JP3455080B2 - Dot code - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to so-called multimedia information including audio information such as voice and music, video information obtained from a camera, a video and the like, digital code data obtained from a personal computer, a word processor and the like. The present invention relates to a dot code which is suitable for recording and / or reproducing, and which is optically readable recorded on a sheet-shaped medium such as paper, various resin films, and metals.

[0002]

2. Description of the Related Art Conventionally, various media such as magnetic tapes and optical disks have been known as media for recording voice, music and the like.

However, even if a large number of copies are made, these media are expensive to some extent, and a large amount of space is required for their storage.

Further, when it is necessary to hand over the voice recording medium to another person in a remote place, whether it is mailed or brought directly, it takes time and time. There was also the problem of this.

Therefore, it has been considered to record voice information on paper in the form of image information which can be transmitted by facsimile and can be reproduced in large quantities at low cost. For example, JP-A-60
As disclosed in Japanese Patent Laid-Open No. 244145, there has been proposed a method in which a small amount of voice is converted into an optical code so that voice information is converted into image information and can be sent by facsimile.

[0006]

However, in the apparatus disclosed in the above publication, the facsimile apparatus is provided with a sensor for reading the optically readable recorded voice, and according to the sensor output. I try to play the sound. Therefore, the optically readable voice information transmitted by facsimile can only be heard at the place where the facsimile device is installed, and the usage such as reproducing the sound by moving the facsimile output paper to another place is assumed. Was not done.

Therefore, if the recording capacity of voice information is increased, it may affect other facsimile transmission / reception, and if the recorded content itself is difficult, while reproducing a large amount of voice. It is possible that you forget the first one. Furthermore, the recording capacity is limited by the recording density and the compression method, and only a voice of about several seconds can be transmitted. Therefore, in order to send a large amount of audio information, the magnetic tape, the optical disk, etc. have to be resorted to.

Further, even for a short time of voice information, since the reproducing apparatus itself is built in the facsimile apparatus, it is inconvenient to repeatedly reproduce the voice information.

In addition to audio information, so-called multimedia information as a whole including video information obtained from a camera, a video, etc., digital code data obtained from a personal computer, a word processor, etc. is inexpensive and has a large capacity. The recording / playback system has not been realized yet.

The present invention has been made in view of the above points, and is a dot code capable of inexpensively recording a large amount of multimedia information including audio information, video information, digital code data, and the like, and repeatedly reproducing it. The purpose is to provide.

[0011]

In order to achieve the above object, the dot code according to the present invention is optically read.
Possible dot code , multiple blocks can be adjacent
Will be placed in ability, each of the blocks, and a non-modulation area including a marker for recognizing the block, the marker structurally distinguishable and to order modulation processing of decorated with each bit value of the data Is composed of a modulation area including a data dot pattern composed of a plurality of dots arranged in dots corresponding to the value, and the non-modulation area further has a block address pattern for indicating the address of the block. It is characterized by including.

That is, according to the dot code of the present invention, each block constituting the dot code is modulated so as to be structurally distinguishable from a non-modulation area including a marker for recognizing the block. Each bit value of the processed data is composed of a modulation area including a data dot pattern composed of a plurality of dots arranged in dots corresponding to the value, and further, in the non-modulation area, the block A block address pattern for indicating an address is arranged.

Therefore, when the data relating to the multimedia information is recorded on the recording medium as a data dot pattern so as to be optically readable, the marker and the block address pattern arranged in the non-modulated area are not modulated, and therefore it is easy. It is possible to identify the array position of each dot of the data dot pattern in the block two-dimensionally by such a marker pair that is easy to recognize, and the positional relationship of each block can be determined from the block address pattern. Since it can be derived, each dot of the data dot pattern can be recorded in a very fine size, and thus inexpensive and large-capacity recording becomes possible.
Further, such a dot code can be repeatedly reproduced.

Alternatively, the dot code according to the present invention is
A optically readable dot code, block
Multiple blocks are arranged so that they can be adjacent to each other, and each of the blocks has been subjected to a modulation process so as to be structurally distinguishable from the non-modulation region including the marker for recognizing the block. Each bit value of data is composed of a modulation area including a data dot pattern composed of a plurality of dots arranged in dots corresponding to the value, and the non-modulation area further indicates an address of the block. Block address pattern and a dot pattern for pattern matching for determining the read position of each dot in the data dot pattern.

That is, according to the dot code of the present invention, each block constituting the dot code is modulated so as to be structurally distinguishable from a non-modulation area including a marker for recognizing the block. Each bit value of the processed data is composed of a modulation area including a data dot pattern composed of a plurality of dots arranged in dots corresponding to the value, and further, in the non-modulation area, the block A block address pattern to indicate the address,
A dot pattern for pattern matching for determining the read position of each dot in the data dot pattern is arranged. Therefore, multimedia information
Optically read the data related to
When recording on a recording medium so that only the
Arranged markers, block address patterns, and patterns
The matching dot pattern is not modulated
The marker pairs that can be easily recognized and are easy to recognize.
Position of the dot pattern for pattern matching
The dot pattern for pattern matching easily
Recognize and use this dot pattern for pattern matching
Arrangement position of each dot of the data dot pattern in the block
It becomes possible to specify the position in two dimensions, and
The lock positional relationship is derived from the block address pattern
It is possible to put out each data dot pattern
It is possible to record dots in a very fine size.
Therefore, it becomes possible to record at a low cost and in a large capacity. Also, this
Dot codes such as can be repeatedly played.

[0016]

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. First, of the multimedia information, embodiments relating to audio information such as voice and music will be described. .

FIG. 1 is a block diagram of an audio information recording apparatus for recording audio information such as voice and music on a paper as an optically readable digital signal according to the first embodiment of the present invention. is there.

The audio signal input by the voice input device 12 such as a microphone or an audio output device is
After being amplified by the preamplifier 14 (AGC is applied in the case of microphone sound), it is converted into a digital signal by the A / D converter 16. The digitized audio signal is subjected to data compression by the compression circuit 18, and then an error correction code is added by the error correction code addition circuit 20.

Thereafter, the memory circuit 22 performs interleaving. This interleaving distributes the data array two-dimensionally according to a predetermined rule, so that when the data is returned to the original array by the reproducing apparatus, paper-like stains and scratches, that is, The error itself is dispersed, which facilitates error correction and data interpolation. This interleaving is performed by appropriately reading and outputting the data stored in the memory 22A by the interleaving circuit 22B.

The output data of the memory circuit 22 is then subjected to a x-address indicating a marker and a two-dimensional address of each block by a data adding circuit 24 in accordance with a predetermined recording format as will be described later in detail. Then, after the y address and the error determination code are added, the modulation circuit 26 performs modulation for recording. Then, after data such as image data to be recorded together with the output data of the audio information is superimposed by the synthesizing circuit 27, the printer system or the plate making system for printing 28 takes measures for printing.

As a result, for example, it is recorded on the paper 30 as a recording medium in a format as shown in FIG. That is, together with the image 32 and the character 34, the sound data converted into a digital signal is printed as the recording data 36. Here, the recording data 36 is composed of a plurality of blocks 38, and each block 38 includes a marker 38A, an error correction code 38B, audio data 38C, x address data 38D, y address data 38E, and an error determination code. It is composed of 38F.

The marker 38A also functions as a synchronizing signal, and uses a pattern that does not normally appear in recording modulation, such as DAT. The error correction code 38B is used for error correction of the audio data 38C. The audio data 38C corresponds to the audio signal input from the voice input device 12 such as the microphone or the audio output device. The x and y address data 38D and 38E are data representing the position of the block 38, and the error determination code 38F is used for error determination of these x and y addresses.

Recording data 36 having such a format
Prints the data of "1" and "0" by the printer system or the printing plate making system 28 by setting "1" with black dots and "0" without black dots as in the case of a bar code, for example. Will be recorded. Hereinafter, such print data will be referred to as a dot code.

FIG. 2B shows a situation where the sound data recorded on the paper 30 as shown in FIG. 2A is being read by the pen-type information reproducing device 40. By tracing over the dot code 36 with the pen-type information reproducing device 40 as shown in the figure, the dot code 36 can be detected, converted into a sound, and heard by the voice output device 42 such as an earphone.

FIG. 3 is a block diagram of the information reproducing apparatus 40 according to the first embodiment of the present invention. In the information reproducing apparatus of the present embodiment, parts other than the audio output device 42 such as headphones and earphones are housed in one portable pen-type housing (not shown). Of course, the speaker may be built in the housing.

The detection unit 44 basically has the same function as that of an image pickup unit such as a television camera. That is, the dot code 36 on the paper surface, which is a subject, is illuminated by the light source 44A, and the reflected light is imaged by the image pickup unit 44D including a semiconductor area sensor or the like via the imaging system 44B such as a lens and the spatial filter 44C. Is detected, and is amplified by the preamplifier 44E and output.

Here, the pixel pitch of the area sensor is set to be less than or equal to the dot pitch of the dot code 36 on the image pickup surface by the sampling theorem. Further, the spatial filter 44C installed on the image pickup surface is also based on this theorem,
It is inserted to prevent the moire phenomenon (aliasing) on the imaging surface. Further, the number of pixels of the area sensor is set in the vertical direction of a predetermined dot code 36 that is defined as being readable at one time in consideration of camera shake when manually scanning the detection unit 44 as shown in FIG. It is set to be larger than the width of.
That is, (A) and (B) of FIG. 4 show the moving state of the imaging area for each certain period when the detecting unit 44 is manually scanned in the direction of the arrow, and in particular, (A) shows dots. FIG. 9B shows a state of manual scanning when the width of the code 36 in the vertical direction is within the imaging area (camera shake is also taken into consideration), and (B) shows that the amount of the dot code 36 is large and the width in the vertical direction is one. The case where the image does not fit in the imaging area is shown. In the latter case,
At the position where the manual scanning of the dot code 36 is started, a manual scanning mark 36A for indicating the dot code 36 is printed.
Therefore, a large number of dot codes 36 can be detected by performing manual scanning a plurality of times along the manual scanning mark 36A.

The image signal detected by the detection unit 44 as described above is then input to the scan conversion and lens distortion correction unit 46. In the scan conversion and lens distortion correction unit 46, the input image signal is first converted into an A / D converter 46A.
Is converted into a digital signal and stored in the frame memory 46B. The frame memory 46B has a gradation of 8 bits.

Further, the marker detection circuit 46C detects the marker 38A by scanning the image information stored in the frame memory 46B as shown in FIG. 4C. θ
The detection circuit 46D detects which address value on the image pickup surface each marker 38A detected by the marker detection circuit 46C corresponds to, and the inclination θ of the image pickup surface with respect to the arrangement direction of the dot codes from the address value. Is calculated. It should be noted that the marker detection circuit 46C is rotated by approximately 90 ° as in the case of FIG. 4C to scan the dots in the scan only in the direction as shown in FIG. 4C, as shown in FIG. When the image of the code 36 is picked up, the inclination θ may not be obtained correctly. That is, when scanning in the lateral direction of the block 38, θ may not be correctly obtained, so the marker detection circuit 46C also performs scanning in the orthogonal direction as shown in FIG. The correct one of the results obtained by the two-direction scanning is selected.

On the other hand, in the lens aberration information memory 46E,
Pre-measured aberration information of the lens used in the image forming system 44B of the detection unit 44 for correcting the distortion of the lens is stored. When the address control circuit 46F next reads the data stored in the frame memory 46B, the value of the inclination θ calculated by the θ detection circuit 46D and the lens aberration stored in the lens aberration information memory 46E are read. The read address according to the information is used as the frame memory 4
6B, and scan conversion in the array direction of data is performed while performing data interpolation by the interpolation circuit 46G.

FIG. 5A shows the principle of data interpolation performed by the interpolation circuit 46G. Basically,
Interpolation data is created by a convolution filter and LPF using pixels around the position Q where data is interpolated. The pixel pitch and the scanning line pitch after the scan conversion are set below the dot pitch of the dot code based on the sampling theorem, as in the case of imaging.

In the case of simple data interpolation using four pixels around the position Q to be interpolated, Q = (D6 × F6)
+ (D7 x F7) + (D10 x F10) + (D11 x F11),
In the case of relatively accurate data interpolation using 16 surrounding pixels, Q = (D1 × F1) + (D2 × F2
) + ... + (D16 × F16) is used to create the interpolation data. Where Dn is the data amplitude value of pixel n, Fn
Is a coefficient of an interpolation convolution filter (LPF) determined according to the distance to the pixel n.

The image of the dot code 36 read from the frame memory 46B after being subjected to the scan conversion as described above is then binarized by the binarization circuit 48 including the latch 48A and the comparator 48B. It The threshold value for the binarization is determined by the threshold value determination circuit 50 by using the value of the histogram for each screen or each block in the screen. That is, the stain on the dot code 36 and the paper 3
The threshold value is determined according to the distortion of 0, the accuracy of the built-in clock, and the like. As the threshold determination circuit 50, it is preferable to use, for example, a circuit using a neural network disclosed in Japanese Patent Application No. 4-131051 by the present applicant.

In parallel with this, the frame memory 46
The image of the dot code 36 read from B is input to the PLL circuit 52, and the clock pulse CK synchronized with the reproduction data is generated. This clock pulse CK is binarized or demodulated after scan conversion, and a data string adjustment unit 56 described later.
Error determination circuit 56A, x, y address detection circuit 56
It is used as a reference clock for B and the memory unit 56C.

The binarized data is demodulated by the demodulation circuit 54, and the error judgment circuit 56A in the data string adjustment unit 56 is provided.
Is input to the x, y address detection circuit 56B. The error determination circuit 56A is configured to operate the error determination code 38 in the block 38.
Using F, it is determined whether or not there is an error in the x, y address data 38D, 38E. If there is no error, the demodulated data from the demodulation circuit is converted to the x, y address detection circuit 5
According to the address detected in 6B, it is recorded in the memory unit 56C for audio data string adjustment. If there is an error, the audio data 38C of the block 38 is not recorded in the audio data string adjustment memory unit 56C.

The purpose of the data string adjusting unit 56 is to determine the accuracy of the scan conversion in the scan conversion and lens distortion correction unit 46 (depending on the accuracy of the reference clock and the S / N of the image sensor), the distortion of the paper, and the like. , To correct a slight shift generated in the data array direction and the scan direction after scan conversion. This will be described with reference to FIG. In the figure, dot codes D1, D2 and D3 indicate data for each block. The pitch of the scan lines 1, 2, 3, ... After scan conversion is
As described above, it may be set to the dot pitch of the data or less based on the sampling theorem, but in FIG.
For completeness, the dot pitch is set to 1/2. Therefore, as is clear from the figure, the dot code D1 can be detected without error in the scan line 3 after scan conversion. And D2
Is detected on the scan line 2 after scan conversion without error, and D3 is similarly detected on the scan line 1 after scan conversion without error.

Then, x in each block 38,
It is stored in the memory unit 56C for adjusting the data string according to the y addresses 38D and 38E.

Next, as shown in FIGS. 4A and 4B, the detection unit 44 is manually scanned, so that the voice dot code 36 on the paper 30 is not leaked and a memory for adjusting the data string is provided. It can be stored in the section 56C.

The voice dot code whose data string has been adjusted by the data string adjusting unit 56 is
According to a reference clock CK 'generated by a reference clock generation circuit 53 which is different from the circuit 52, the data is read from the memory unit 56C for adjusting the data string. Then, at this time, de-interleaving is applied by the de-interleaving circuit 58,
Converted into a formal data string. Next, the error correction circuit 60 performs error correction using the error correction code 38B in the block 38. Then, the decoding circuit 62 decodes the compressed data, and the data interpolation circuit 64 further interpolates the audio data whose error cannot be corrected. After that, it is converted into an analog audio signal by the D / A conversion circuit 66, amplified by the amplifier 68, and converted into sound by the voice output device (earphone, headphone, speaker, etc.) 42.

As described above, audio information such as voice and music can be recorded on paper, and the reproducing device is a small portable device so that a printout or a facsimile transmission thereof can be performed. , Or printed in the form of a book by printing plate making, anywhere
You will be able to listen again and again.

The data string adjusting memory unit 56C in the data string adjusting unit 56 is not limited to the semiconductor memory, and
Floppy disks, optical disks, magneto-optical disk, it is possible to use other storage media like.

Various applications are conceivable as applications of the audio information recorded as described above. For example, for general use, language textbooks, sheet music, various texts such as distance learning, product specifications, manuals for repairs, dictionaries in foreign languages, encyclopedias, books such as picture books, product catalogs, travel guides, direct mail and guidance. Letter, newspaper, magazine, leaflet,
Albums, congratulations, postcards, etc. can be considered. For business use, FAX (voice & fax) business instruction,
Minutes, electronic blackboards, OHPs, identification cards (voiceprints), business cards, telephone memos, sticky notes, supply products (consumables) in the form of rolls of fine paper, and the like are conceivable. Here, as shown in FIG. 5 (B), the consumable item means that a double-sided tape or a sticky paper such as a sticky note is provided on the back surface of the rolled paper 30A, Record the dot code 36 on, and separate it as needed.
It is designed to be attached to various objects (hereinafter referred to as reel seal). Further, as shown in (C) of the figure, the width of the paper 30A is widened so that a plurality of dot codes 36 can be recorded, and the manual scanning mark 36B is used as a guideline for the manual scanning of the detection unit 44. May be printed vertically and horizontally. At the same time, the mark 36B can be used as a guide for the recording position of the dot code 36. That is, if a sensor is provided in the printer system 28 and the mark 36B is read by the sensor to find the beginning of printing, the dot code 36 is always printed in the area surrounded by the mark 36B. Since it is possible to perform the manual scanning along the mark 36B, the recorded audio information can be surely reproduced. Of course, the mark 36B may be printed when the dot code 36 is printed.

The recording time of audio information is 20
In the case of a general facsimile of 0 dpi, for example, 1 inch × 7 inch (2.54 cm × 1) along one side of the paper.
When data is recorded in an area of 7.78 cm, the total number of data is 280 kbit. From now on, a marker, an address signal, an error correction code, an error determination code (however, the error determination code in this case is the x, y address 38D, 38
In addition to E, audio data 38C is also subjected to error determination) (30%) is subtracted, 196 kbit
become. Therefore, the recording time when the voice is compressed to 7 kbit / s (mobile communication bit rate) is 28 seconds. When recording on the entire back side of A4 size double-sided facsimile paper, 7 inches x 10 inches (17.78 cm x 2)
Since the area of 5.4 cm) can be taken, it is possible to record 4.7 minutes of voice.

Further, in the case of a 400 dpi G4 facsimile, as a result of the same calculation as above, 7 inches × 10
It is possible to record 18.8 minutes of voice in the inch area.

For high-quality printing of 1500 dpi, 5 mm
When printed in an area of × 30 mm, the result of the same calculation as above results in a voice recording of 52.3 seconds. Also,
When printed on a 10 mm x 75 mm tape-shaped area, high-quality sound (30 kbi after compression is possible) that enables music
When calculated with an audio signal of t / s), one minute of audio recording is possible.

FIG. 7 is a diagram showing the configuration of the second embodiment of the present invention. The second embodiment is an example in which an xy address type image pickup unit such as a memory and a randomly accessible CMD is used as an image pickup device, and the detection unit 44 and the scan conversion and lens distortion correction circuit 46 of the reproducing apparatus are used.
Only the difference from the first embodiment. That is,
The detection unit and the scan conversion unit 70 are the xy address type image pickup unit 7
As in the first embodiment, the image pickup data stored in the memory 0A is subjected to marker detection, and four pieces of surrounding data to be interpolated when the marker data is read are provided to the decoder address generators 70B and x ,.
The y decoders 70C and 70D sequentially read the data and input it to the interpolation unit 72. In the interpolator 72, the coefficient is sequentially read from the coefficient generation circuit 70E with respect to the input data, multiplied by the multiplier 70F, and further converted into an analog cumulative addition circuit including an adder 70G, a sample and hold circuit 70H, and a switch 70I. Are cumulatively added, and sample-and-hold is performed by the sample-and-hold circuit 70J, and the scan-converted dot code is converted into the binarization circuit 4 described above.
8, the threshold value determination circuit 50, and the PLL circuit 52.

With this structure, the same function as that of the first embodiment can be achieved, the frame memory 46 can be eliminated, and the cost and size can be reduced. . Furthermore, by further incorporating the xy address type image pickup unit 70A, the address generation unit 70B, the decoders 70C and 70D, and the interpolation unit 72 into a single substrate to form an IC, further miniaturization can be achieved.

FIG. 8 is a diagram showing the configuration of the third embodiment of the present invention. In this embodiment, the dot code 36 is recorded on the paper 30 on which a picture or a character is printed with a transparent paint (ink) 74 that is easily specularly reflected (totally reflected). Then, in the detection unit 44, the light source 44A and the imaging system 44
Polarizing filters 44F and 44G are provided between B, and the polarization planes of these polarizing filters 44F and 44G are aligned, so that the reflected light from the inside (the surface of the paper 30) and the transparent paint 74 escape according to the code. The polarization direction of the reflected light from the place where the hole 74A is opened is scattered, and the polarizing filter 44G cuts half of it.
Further, since the difference in light amount between the normal reflected light and the total reflected light is originally large, the contrast of the dot code recorded with the transparent paint 74 is emphasized and an image is captured.

Further, the paper 30 is subjected to a surface treatment such as a mirror finish so that the surface is easily specularly reflected, and the transparent coating material 74 is made of a material having a refractive index higher than that of the surface-treated surface and ¼λ. If a film with a certain thickness (in consideration of the change in the optical path length depending on the incident angle, the optical path length in the transparent paint becomes 1/4) is used, the effect of the reflection-amplifying coat prevents the light that obliquely strikes. , More easily amplified and easily reflected on the surface (regular reflection).

In this case, for example, the dot code is formed by
Fine chemical etching or the like is performed to roughen the holes corresponding to the dots to reduce the reflectance.

When the dot code 36 is recorded by the transparent paint 74 in this way, it can be recorded on a character or a picture. Therefore, when the dot code 36 is used together with the character or the picture, the recording is different from that in the first embodiment. The capacity can be increased.

Also, instead of the transparent paint, a transparent fluorescent paint may be used, or it may be colored and multiplexed. In the case of this color, a normal color ink can be used, or a transparent ink can be mixed with a dye to form a color.

Here, as an example, the transparent ink may be an ink composed of a volatile liquid and a binder (for example, there are phenol resin varnish, linseed oil varnish, and alkyd resin), and the dye may be a pigment.

Next, a portable voice recorder to which the audio information recording device is applied will be described. 9A and 9B are external views thereof. This portable voice recorder includes a main body 76, a main body side and a voice input section side attaching / detaching member (a surface fastener, a magic tape (registered trademark) ) 78A,
A voice input section 80 that can be detachably attached to the main body 76 by 78B
Consists of. A recording start button 82 and a print sheet discharging portion 84 are provided on the surface of the main body 76. The main body 76 and the voice input section 80 are connected by a cable 86. Of course, the signal may be transmitted from the voice input unit 80 to the main body 76 by wireless or infrared rays.

FIG. 10 is a block diagram of such a portable voice recorder. The voice input from the microphone 88 is amplified by a preamplifier 90, converted into a digital signal by an A / D converter 92, and then compressed by a compression processing unit (ADPC).
M) 94. The compressed data is
The error correction code adding unit 96 adds the error correction code, the result is supplied to the interleaving unit 98, each data is stored, and then the interleaving process is performed. The data interleaved in this way is further processed by the address data addition unit 100 to determine the address of the block and the error determination code (CRC) for the address.
Etc.) is added, and the result is input to the modulation circuit 102. In this modulation circuit 102, 8-bit data, such as 8-10 modulation, is converted into data having a different bit number of 10 bits. After that, in the marker adding unit 104,
A marker is generated and added using a data string that does not exist in the 256 data strings associated by the modulation circuit 102.

The data to which the marker has been added in this way is sent to the simple printer system 106, and is sent to the printer shown in FIG.
And, as shown in (B), the reel seal 108 is printed, and the sheet is discharged from the print sheet discharging section 84. in this case,
The simple printer system 106 prints the date and time measured by the timer 110 on the reel seal 108.

Note that each of the above parts has a recording start button 82.
It is controlled by the control unit 112 according to the operation. Also,
Of the above-mentioned units, the distance from the microphone 88 to the voice input unit 80 is not particularly limited. For example, here, the voice input unit 80 includes the microphone 88, the preamplifier 90, and the A / D converter. 92 shall be incorporated.

FIG. 12 is an operation flowchart of the portable voice recorder having such a configuration. That is, the main body 76
When the recording start button 82 provided at the step S12 is pressed (step S12), while the button is pressed (step S1).
4) The processing from the voice input to the dot code 114 printing processing on the reel seal 108 is performed (step S1).
6). Then, when the pressing of the recording start button 82 is stopped, the recording start button 8 is again pressed within a predetermined time.
2 is pressed (step S18),
If it is determined that the button has been pressed, the process returns to step S14 and the above process is repeated. However, if the recording start button 82 is not pressed within the fixed time, the timer 110 refers to the current date and time (step S
20), the reel seal 108 is printed on the blank portion 116 while feeding the referred date and time (step S22).

In such a portable voice recorder, as shown in FIG. 9A, when the main body 76 and the voice input section 80 are connected, the user holds the main body 76 by hand and operates the voice input section 80. Dot code 114 by moving the voice closer to your mouth
Is recorded on the reel seal 108. Further, as shown in FIG. 9B, the main body 76 and the voice input section 80 are separated, and the voice input section 80 is attached to the handset side of the handset of the telephone by using the attaching / detaching member 78B. Instead of making a note of the contents, the other party's message can be directly recorded on the reel seal 108 as the dot code 114. Moreover, in this case, as shown in FIGS. 11A and 11B, not only the date and time are printed on the reel seal 108, but also the margin portion 116 is formed. You can write a comment, such as who or to whom.

The voice input section 80 may have various modes other than the structure in which it is attached to and detached from the main body by the attachment / detachment member as described above. For example, as shown in FIGS. 11C and 11D, an earphone type can be used. In the case of such an earphone type voice input unit 80, as shown in FIG.
By pulling it out of the voice input section storage section 118 of No. 6 and inserting it into the user's ear, it becomes possible to record it in the form of dot code while listening to the voice of the other party heard from the handset side of the handset of the telephone.

Further, in the above description, the recording start button 82
The dot code is printed only while the button is kept pressed. However, a recording end button is separately provided on the main body 76, and the dot code is printed from when the recording start button 82 is pressed once until the recording end button is pressed. Printing may be performed.

The recording device may be a recording / reproducing device by incorporating the reproducing function as shown in FIG. In that case, the earphone type voice input unit 80 may also have the function of an earphone.

In the above embodiments, audio information such as voice and music has been described as an example of the information to be recorded. However, the information to be recorded is not limited to audio information but can be obtained from a camera, a video or the like. Of handling so-called multimedia information including video information to be stored, digital code data such as text data obtained from a personal computer (hereinafter, referred to as a personal computer), a word processor (hereinafter, referred to as a word processor), and the like explain.

FIG. 13 is a block diagram of a multimedia information recording apparatus for recording such multimedia information.

Of the multimedia information, audio information is input from the microphone or audio output device 120 as in the case of FIG.
After being amplified by, the signal is converted into a digital signal by the A / D converter 124 and supplied to the compression processing unit 126.

In the compression processing unit 126, the input digital audio signal is selectively supplied to the voice compression circuit 130 such as the ADPCM circuit and the voice synthesis coding circuit 132 by the switch 128. The voice compression circuit 130 performs data compression by performing adaptive differential PCM on the input digital audio information. The voice synthesis coding circuit 132 recognizes one voice from the input digital audio information and then converts it into a code. This is because the ADPCM encodes it in the form of voice information to reduce the amount of data, that is, the raw data is processed, whereas it is relatively changed by once changing to another synthesized code. It reduces the amount of data.
Regarding the switching of the switch 128, for example, the user manually switches the switch 128 according to the purpose. Alternatively, for example, high-quality information such as information from an audio output device is passed through the voice compression circuit 130, and, for example, a person's voice or comment from a microphone is passed through the voice synthesis coding circuit 132. By predetermining in this way, it is also possible to adopt a configuration in which which audio information is input is recognized at the front stage of the switch and automatically switched.

Further, various data which has already been formed as digital code data from a personal computer, word processor, CAD, electronic notebook, communication or the like is first transmitted through an interface (hereinafter referred to as I / F) 134 in a data form. It is input to the determination circuit 136. The data form determination circuit 136 basically determines whether or not compression is possible in the compression processing unit 126 in the subsequent stage, and some compression processing has already been performed on the data, and the compression processing unit 126 in the subsequent stage. For information for which the effect of the above is not obtained, the compression processing unit 126
Is directly passed to the subsequent stage of the compression processing unit 126, and when the input data is uncompressed data,
It is sent to the compression processing unit 126.

The data judged to be uncompressed code data by the data form judging section 136 is inputted to the compression processing section 126, and the code data is optimized by the compression circuit 138 such as Huffman, arithmetic code and dibrempel. The compression process is performed to compress to. The compression circuit 138 also performs compression processing on the output of the speech synthesis coding circuit 132.

The voice synthesis coding circuit 132 is used.
May recognize character information other than voice and convert it into voice synthesis code.

The image information of the camera or video output device 140 is supplied to the compression processing unit 126 after being amplified by the preamplifier 142 and A / D converted by the A / D converter 144.

In the compression processing unit 126, the image area determination and separation circuit 146 determines whether the input image information is a binary image such as a handwritten character or a graph or a multi-valued image such as a natural image. . This image area determination and separation circuit 146
For example, the binary image data and the multi-valued image data are separated by using a discriminant image area separation method using a neural network as disclosed in Japanese Patent Application No. 5-163635 of the present applicant. Then, the binary image data is compressed by a general binary compression processing circuit 148 such as MR / MH / MMR such as JBIG as binary compression, and the multivalued image data is a still image such as DPCM or JPEG. It is compressed by the multi-level compression processing circuit 150 using the compression function of.

The data subjected to the compression processing as described above is appropriately combined by the data combining processing section 152.

It should be noted that it is not always necessary to provide all of the respective information input and compression processing systems in parallel, and one or a plurality of systems may be combined as appropriate according to the purpose. Therefore, the data synthesizing processing unit 152 is not always necessary, and if the data system has only one type, omit it.
The configuration may be such that it is directly input to the error correction code addition unit 154 in the next stage.

The error correction code adding section 154 adds the error correction code and inputs it to the data memory section 156. The data memory unit 156 stores the respective data, and thereafter, interleave processing is performed. This actually reduces the error when recorded as a dot code and played it back, eg
This is a process of dispersing continuous data strings at appropriately distant positions in order to improve the correction capability by reducing block errors due to noise or the like. That is, the work of reducing the risk of burst errors into bit errors is performed.

With respect to the interleaved data, the address data adding unit 158 further causes the address of the block and the error determination code (CR) for the address.
C, etc.) is added, and the result is input to the modulation circuit 160. The modulation circuit 160 uses, for example, 8-10 modulation.

In the above embodiment, it goes without saying that a code for error correction may be added after interleaving.

After that, the marker adding section 162 generates and adds a marker by using a data string that does not exist in the 256 data strings associated by the modulation circuit 160. By adding the marker after the modulation as described above, it is possible to eliminate the problem that even the marker is also modulated and it is difficult to recognize the marker as a marker.

The data thus added with the marker is sent to the synthesizing / editing processing section 164 and recorded on a recording paper other than the generated data, for example, synthesized with an image, a title, characters, or the like, or The layout and the like are edited, and the output format to the printer and the data format corresponding to the printing plate making are converted and sent to the next printer system or the printing plate making system 166. Then, the printer system or the printing plate making system 166 finally prints on a sheet, a tape, a printed matter, or the like.

The editing process in the synthesizing / editing processing unit 164 is performed in code units in word units, content breaks, etc., in order to match the paper surface information and dot code layout, the dot size of the code to the resolution of the printing machine, printer, etc. This includes editing work such that the length is appropriately divided and the stage is changed, that is, the line is changed to the next line.

The printed matter thus printed is, for example, F
It is transmitted by AX168. Of course, instead of printing the data generated by the composition / editing processing unit 164, the data may be directly faxed.

Here, the concept of the dot code 170 in the present embodiment will be described with reference to FIG. In the data format of the dot code 170 of the present embodiment, one block 172 includes a marker 174, a block address 176, address error detection / correction data 178, and a data area 1 in which actual data is stored.
It consists of 80 and. That is, in the embodiment described with reference to FIG. 2A, one block is configured one-dimensionally in the line direction, but in the present embodiment, it is developed two-dimensionally. It is formed in a shaped shape. The blocks 172 are arranged vertically, horizontally, and two-dimensionally, and they are gathered to form a dot code 170.

Next, the structure of the multimedia information reproducing apparatus will be described with reference to the block diagram of FIG. This information reproducing apparatus is a scanning unit for recognizing the image data supplied from the detection unit 184 for reading the dot code from the sheet 182 as a recording medium on which the dot code 170 is printed and the detection unit 184 as a dot code for normalization. A conversion unit 186, a binarization processing unit 188 for converting multi-valued data into a binary value, a demodulation unit 190, an adjustment unit 192 for adjusting a data string, a data error correction unit 194 for correcting a read error and a data error during reproduction, and a data A data separation unit 196 for separating the data according to each attribute, a decompression processing unit for data compression processing according to each attribute, a display unit or a reproduction unit, or another input device.

In the detection section 184, the dot code 170 on the sheet 182 is illuminated by the light source 198, and the reflected light is provided with the imaging optical system 200 such as a lens and the spatial filter 202 for removing moire and the like. The image information is converted into an electric signal through an image pickup unit 204 such as a CCD or CMD, which is detected as an image signal, amplified by a preamplifier 206, and output. The light source 198, the imaging optical system 200, the spatial filter 202, the imaging unit 204, and the preamplifier 20.
Reference numeral 6 denotes an external light shielding portion 208 for preventing external light disturbance.
Composed within. The image signal amplified by the preamplifier 206 is converted into digital information by the A / D conversion unit 210 and supplied to the scan conversion unit 186 at the next stage.

The image pickup section 204 is controlled by the image pickup section control section 212. For example, when an interline transfer type CCD is used as the imaging unit 204,
The image pickup unit control unit 212 stores a V blank signal for vertical synchronization, an image pickup device reset pulse signal for resetting information charges, and a charge transfer accumulation unit arranged two-dimensionally as control signals for the image pickup unit 204. Charge transfer gate pulse signal for transmitting the electric charge to a plurality of vertical shift registers, and a horizontal charge transfer CLK which is a transfer clock signal of the horizontal shift register for transferring the electric charge in the horizontal direction and outputting it to the outside.
A signal, a vertical charge transfer pulse signal for vertically transferring the plurality of vertical shift register charges and sending the charges to the horizontal shift register, and the like are output. The timing of these signals is shown in FIG.

Then, the image pickup section control section 212 gives the light source a light emitting cell control pulse for timing the light emission of the light source 198 in synchronization with this timing.

Basically, the timing chart of FIG. 16 is a conceptual diagram for one field. The image data is read from the V blank of this one field to the V blank. The light source 198 performs pulse lighting instead of continuous lighting, and while synchronizing in field units,
Subsequent pulse lighting shall be performed. In this case, the timing is controlled such that the exposure is performed during the V blanking period, that is, while the image charge is not being output, so that the clock noise for pulse lighting does not enter the signal output. That is, since the light emitting cell control pulse is a very thin digital clock pulse that is generated instantaneously and gives a large amount of power to the light source, it is possible to prevent noise due to it from entering the analog image signal. This is necessary, and as a measure therefor, the light source is pulsed during the V blanking period. By doing so, the S / N is improved.
Further, pulse lighting means shortening the light emission time, and thus has a great effect of eliminating the influence of blurring due to shake and movement of manual operation. This enables high-speed scanning.

In addition, in order to minimize the S / N deterioration even when the reproducing apparatus is tilted and the disturbance such as the outside light enters for some reason despite the existence of the outside light shielding portion 208. , Immediately before the light source 198 is caused to emit light in the V blanking period, the image sensor reset pulse is output once to reset the image signal, light is emitted immediately after that, and immediately thereafter.
I try to read it.

Now, returning to FIG. 15, the scan conversion unit 186 is executed.
Will be explained. The scan conversion unit 186 recognizes the image data supplied from the detection unit 184 as a dot code,
This is the part for normalizing. As the method, first, the image data from the detection unit 184 is stored in the image memory 214, and once read out from the image memory 214 and sent to the marker detection unit 216. The marker detection unit 216 detects a marker for each block. Then, the data array direction detection unit 218
Uses the marker to detect rotation or tilt and the data array direction. Based on the result, the address controller 220 reads out the image data from the image memory 214 so as to correct it and supplies it to the interpolation circuit 222.
At this time, lens aberration information is read from the memory 224 for correcting distortion of lens aberration in the imaging optical system 200 of the detection unit 184, and lens correction is also performed. Then, the interpolation circuit 222 performs interpolation processing on the image data and converts it into the original dot code pattern.

The output of the interpolation circuit 222 is the binarization processing unit 1
88. Basically, as can be seen from FIG. 14, the dot code 170 is a white and black pattern, that is, binary information, and is thus binarized by the binarization processing unit 188. At that time, the threshold determination circuit 226 adaptively performs binarization while determining the threshold in consideration of the influence of disturbance, the influence of the signal amplitude, and the like.

Since the modulation described with reference to FIG. 13 is performed at the time of recording, the demodulation section 190 first demodulates it, and then the data is input to the data string adjustment section 192.

In the data string adjusting unit 192, the block address detecting unit 228 first detects the block address of the above-mentioned two-dimensional block, and then the block address error detecting / correcting unit 230 detects the block address error. After that, the address control unit 23
In 2, the data is stored in the data memory unit 234 in units of blocks. By storing data in block address units in this way, data can be stored without waste even if data is skipped midway or entered midway.

After that, the data read from the data memory unit 234 is corrected by the data error correction unit 194. The output of the error correction unit 194 is branched into two, and one of them is sent as digital data to a personal computer, a word processor, an electronic notebook, etc. via the I / F 236. The other is supplied to the data separation unit 196, where images, handwritten characters and graphs, characters and line drawings,
Sound (2: the original sound and the one with speech synthesis)
Types).

The image corresponds to a natural image and is a multivalued image. The decompression processing unit 238 performs decompression processing corresponding to JPEG at the time of compression, and the data interpolation circuit 240 further interpolates error-uncorrectable data.

For binary image information such as handwritten characters and graphs, the decompression processing unit 242 performs decompression processing for MR / MH / MMR and the like performed by compression, and further the data interpolation circuit 244. Interpolation of data that cannot be error-corrected is performed.

Characters and line drawings are converted into another pattern for display through a PDL (page description language) processing unit 246. In this case, as for the line drawing and the characters, if the compression processing for the code is performed after being encoded, the expansion processing unit 248 corresponding thereto performs the expansion (Huffman, Gibblempel, etc.) processing, and then PD
It is adapted to be supplied to the L processing unit 246.

The data interpolation circuits 240, 244 and P
The output of the DL processing unit 246 is the synthesis or switching circuit 25.
The data is combined or selected by 0, converted into an analog signal by the D / A converter 252, and then displayed on a display device 254 such as a CRT (television monitor) or FMD (face mounted display). The FMD is a spectacles-type monitor (handy monitor) for wearing on the face, and is effective for applications such as virtual reality, and for viewing a large screen in a small place.

As for the voice information, the decompression processing unit 2
At 56, decompression processing is performed on the ADPCM, and at the data interpolating circuit 258, error-uncorrectable data is interpolated. Alternatively, in the case of voice synthesis, the voice synthesis unit 260 receives the voice synthesis code, synthesizes the voice from the code, and outputs the synthesized voice. In this case,
When the code itself is compressed, the decompression processing unit 262 performs decompression processing such as Huffman or Jiblempel in the same manner as the above character and line drawing, and then performs voice synthesis.

Further, as shown in FIG. 17, the character information may be output as voice information by the voice synthesizing unit 260 after the sentence recognition unit 271 recognizes the sentence.

The decompression processing section 262 can also be used as the decompression processing unit 248. In this case, the data is expanded by the switches SW1 and SW1 according to the attribute of the data to be decompressed.
2, SW3 is appropriately switched, and the PDL processing unit 24
6, or input to the voice synthesizer 260.

The data interpolation circuit 258 and the voice synthesis unit 26.
The output of 0 is synthesized or selected by the synthesis or switching circuit 264, converted into an analog signal by the D / A conversion unit 266, and then output to a speaker, headphones, or other similar audio output device 268.

Characters and line drawings are directly output from the data separating unit 196 to the page printer, plotter 270, etc., and the characters are printed on paper as word processing characters, or line drawings are output as a plotter as drawings. It can also be done.

Of course, for images, CRT and FM
Not only D, it is also possible to print with a video printer or the like, and it is also possible to take a picture of the image.

Next, the data string adjusting unit 192 will be described. Here, in order to be applied to the above-described audio information reproducing apparatus (see FIG. 3), the dot code is a block address 272A indicated by reference numeral 272 and its error correction data 272B as shown in FIG.
The blocks provided in the first line are arranged two-dimensionally, and a line-shaped marker 274 as shown in FIG.
Will be described in the vertical direction, and a line address 276A indicated by reference numeral 276 and error detection data 276B will be arranged for each line of each block.

In the present embodiment, as compared with the scanning method described with reference to FIG. 6, as shown in FIG. 18 (C), the pitch is doubled for each line, and the center of the marker is further reduced. After detecting, the center line of the marker is equally divided by twice the number of dots. That is, as shown in FIG.
In the first scan, the dots 278 are finely
Width of 1/2, that is, 1/4 is taken in. In that case, the pitch is the same as that of the dots 278.
Therefore, the data is taken every other dot. In this way, data up to the CRC error detection data 276B, for example, if one block has 64 dots, 64 dots are fetched every other dot.

First, it is confirmed whether or not the line address can be actually read by using the latter line address 276A and the CRC error detection data 276B for the line address. If this line address can be read, it is determined that the data dot before that line has also been correctly read. If it is judged to be wrong, the dot is shifted to the right, for example, to the right, and the second scanning is performed (black circle in (D) of the figure). This is 6
Capture all 4 dots and check if the line address was actually read in the same manner. 1 if wrong
3rd scan, moving 1 dot down from the 3rd dot,
If it is still wrong, move it 1 dot to the right and move it to 4
Perform the second scan.

As described above, if scanning of one line is repeated four times, it seems that the data can be correctly read at least once, so that when it is determined that the data can be read correctly, the data is written to the data memory unit 234. .

In this case, when the line address of the fetched line is recognized as, for example, "0" (start address), that is, the first data, it is determined that the previous data is the block address 272A and the error correction data 272B. . It should be noted that the error correction data 272B may be, for example, CRC for detecting an error in the block address, or may be added to the error correction depending on the purpose, and may also be used as the Reed-Solomon error correction for the block address.
Then, when the first address line 0 is recognized, the block address 272A is read first, and it is determined from this address data what number block the block is. On the other hand, since the actual data is entered from the next line, they are read and the data is written in the block of the data memory unit 234 corresponding to the block.

In the above description, if there is no error while scanning one line, it is assumed that the scanning of the next line is skipped. However, the scanning should be repeated four times for each line. You can At that time, it is determined that the error does not occur multiple times, but the data memory unit 234 stores
Since the same data is only written at the same address, there is no problem. To simplify the process, scanning is repeated 4 times. When speed is prioritized, the former scanning method is adopted.

The actual configuration of the block address detection unit 228 and the block address error detection / correction unit 230 for realizing the above operation of the data string adjustment unit 192 will be described with reference to FIG.

When the binarized interpolation data enters the shift register 190A for 10 bits, the demodulation section 190 converts it into 8 bits by a look-up table (LUT) 190B.

In the data string adjusting unit 192, the demodulated data is temporarily stored in the buffer memory (all 64 dots are filled) under the control of the write address control unit 280.
It is stored in 82. Then, the data read address control unit 284 reads only the line address information and the CRC information for the address, and the line address error detection circuit 286 performs error detection. When the determination signal indicating the result of this error detection is true, that is, there is no error, the data read address control unit 284 reads the information before the line address information, that is, the actual data information from the buffer memory 282.

On the other hand, the start address detection circuit 288
Confirms whether the line address for which an error has been detected by the line address error detection circuit 286 is a start address. When the start address is detected, the start address detection circuit 288 informs the block address detection circuit 290 that the line has a block address as information, and accordingly, the block address detection circuit 290 causes the block address detection circuit 290 to change the buffer memory. 282
A block address is detected from the data read from the error detection circuit 292, and the error detection circuit 292 performs error detection and correction. Then, the result is latched by the address control unit 232 of the data memory unit 234 as a block address.

Note that only error detection is added to the line address in order to obtain an accurate read position, but the block address is used as address information, so an error correction code is added.

Since the subsequent lines thereafter become data lines sequentially, they are written as data in the data memory section 234. At that time, depending on the processing, the line address is also output if necessary. Alternatively, if a counter is provided internally, the line address can be automatically incremented internally.

Then, when the next start address "0" is detected, it is recognized as the next block, and the same operation is repeated for all the blocks.

On the other hand, the line address error detection circuit 28
The determination signal output from 6 is also supplied to the address control unit 220 of the image memory 214. This is a signal necessary for jumping to the next line when the data becomes true in order to reduce the time in the four scannings per line.

In the above example, the line address error detection circuit 286 performs address detection for the interpolation data by using the same address information four times until it becomes true. Then, when the data becomes true, the address is once skipped to the data line of the dot next to the new next line to create the interpolated data, and then four points in the interpolated data are read out. Therefore, for such control, a determination signal is passed to the address control unit 220 of the image memory 214, and the same address is generated four times to be interpolated, or the order of interpolation is changed and read, or The address is rewritten to the next line, the data on that line is output, and the data is output four times while being interpolated.

Although not shown in the figure, the address control section 232 of the data memory section 234 performs mapping to the data memory section 234. When further reading, the address control section 232 controls de-interleaving. Also do. Again, using a lookup table etc., for example, when an address for each dot occurs,
The data obtained by combining the block, the line, and the dot address is converted into a memory data string that actually appears in a lookup table using a ROM or the like. This is the work of de-interleaving (de-shuffling), and the data is read out in the form of the original data string only after the processing is performed. Of course, this de-interleaving may be performed at the time of reading from the data memory unit 234, or at the time of writing, such conversion is performed once and data is written (mapped) in such an order that the data is scattered.
You can do so.

Further, in this example, the marker 274 has a line shape, but it may be a circle as shown in FIG. 14 or a square marker. Once the marker is detected, the blocks are read on a line, so the marker does not necessarily have to be linear. For example, as shown in (A) to (C) of FIG. 20, markers 294, 29 of circles, squares, and rectangles are used.
6,298 are conceivable.

If the printed code has no partial bleeding or deviation and is almost precise, it can be said that (approximately center = accurate center), and therefore accurate center detection described later is omitted. It can be processed only by the approximate center detection process described later. However, in this case, in order to detect the arrangement direction, dots 294A for detecting the arrangement direction,
296A and 298A are provided.

FIG. 20D shows another mode of the reproducing apparatus for multimedia information. This is the detection unit 18
No. 4 A / D converter 210 is moved to the scan converter 186, and the block address detector 22 of the data string adjuster 192 is moved.
8 and the block address error detection / correction unit 230 is performed in the scan conversion unit 186.
Since the data error correction unit 194 and the subsequent components have the same configuration as that of FIG. 15, they are omitted in the figure.

That is, in FIG. 20D, the biggest difference from the configuration shown in FIG.
6 and the data string adjusting unit 192. In this embodiment, the function of the data string adjustment unit 192 is changed to the scan conversion unit 186.
It is assumed that the processes from the marker detection unit 216 to the address control unit 220 are performed simultaneously. That is, the marker detection unit 216 detects the marker, and the data array direction detection unit 218 detects the data array direction, that is, the tilt, rotation, and direction. Then, the block address detection, error determination, and accurate center detection unit 300 detects the block address and makes the error determination, and detects the correct center, that is, the true center, depending on whether it is wrong or not. In this case, since the block address is detected in detecting the true center thereof, after the marker and block address are interpolated by the next marker and block address interpolating unit 302, the information of the block address is converted into data. The address control unit 232 of the memory unit 234 is also provided.

Further, similar to the configuration of FIG. 15, address control is performed by the address control unit 220 based on the data of the interpolation processing of the block address to control the address, the writing and the output to the image memory 214. To do.

Other than that, there is no functional difference from the embodiment of FIG.

Note that in FIG. 15 and FIG. 20 (D), the A / D conversion unit 210 in the detection unit 184 converts the data into, for example, 8-bit multivalued digital data, and the subsequent processing is performed. , A / D conversion unit 210, and a binarization processing unit (comparator) 188 and threshold value determination circuit 22.
6 may be arranged at the A / D conversion unit 210 and all the subsequent processing may be performed with binary data.

In this case, the interpolation circuit 222 uses the pixel data around the interpolation address coordinates obtained from the address control unit 220 as shown in FIG. 5A to perform 4-point or 16-point interpolation. Instead of so-called interpolation processing, pixel data closest to (neighboring) the interpolation address coordinates can be adopted as data.

By performing binarization and processing instead of A / D conversion, for example, 1 is obtained as compared with the case of 8 bits.
The number of signal lines is / 8, and the data amount. Therefore, the memory capacities of the image memory 214 and the data memory unit 234 are also reduced to ⅛, and the processing of each unit is simplified. The circuit scale is greatly reduced, the processing amount is greatly reduced, and the processing time is significantly reduced. This will contribute to downsizing, cost reduction and speedup of the device.

The address output of the address controller 220 is the interpolation circuit 2 in the case of FIG. 15 and FIG.
At the time of outputting the image data to 22, the pixel address becomes, for example, four points around the interpolation address coordinate, and the interpolation circuit 222
The distance information is used for calculating a weighting coefficient for each pixel address by a signal line (not shown). Alternatively, each pixel address and the interpolation address coordinate data may be sent, and the interpolation circuit 222 may obtain the distance from each pixel address to obtain the weighting coefficient.

When processing binary data as described above, the address controller 220 outputs a pixel address near the interpolation address coordinate. Therefore, in this case, the data output from the image memory 214 is directly input to the demodulation unit 190.

Now, a specific example of the dot code shown in the conceptual diagram of FIG. 14 will be described with reference to FIGS.

As shown in the conceptual diagram of FIG. 14, the blocks 304 are two-dimensionally arranged, and block addresses 306 are added to them. The block address 306 has an address corresponding to the X address and the Y address. For example, in FIG. 21A, the upper left block is (X address, Y address) = (1,
1). On the other hand, the block address of the block on the right side is (2, 1), and in the same manner, the X address is incremented toward the right, and the Y address is incremented toward the bottom. So, block address 3 in all blocks 304
06 is added.

Here, the lowermost marker and the rightmost marker are dummy markers 308. That is, the block 304 for a certain marker 310 is the data diagonally to the lower right surrounded by the four markers 310 including it, and the markers at the bottom and the right are the second from the bottom and the second from the right. Auxiliary markers, or dummy markers 308, arranged to define blocks for the markers
Is.

Next, the contents of the block 304 will be described. As shown in FIG. 21B, the block 304
The block address 306 and the error detection code 312 of the block address are added between the marker 310 and the lower marker. Similarly, the block address 306 and its error detection code 312 are added between the marker 310 and the right marker. In the conceptual diagram of FIG. 14, there is a marker at the upper left of the block and the block address is arranged at the lower right, but in the present embodiment, the block address 306 is arranged at the left side and the upper side, and the marker 310 is arranged.
Is placed in the upper left corner. Although the block address 306 is recorded in two locations in one block, it may be recorded in one location. But,
By recording in two places, even if noise occurs in one block address and an error occurs, it can be detected reliably by detecting the other address, so it is better to record in two places. preferable.

The position of the block data with respect to a certain marker, the position of the block address thereof, the position of the dummy marker on the code determined by the position, etc. are not limited to those in the preceding example.

Next, a pattern example of the marker 310 will be described. As shown in FIG. 20C, in the present embodiment,
As the marker 310, a circular black pattern 310A having a diameter of 7 dots is adopted. And that black circle 31
The portion 310B around 0A is white so that the black portion of the marker can be easily identified. Reference numeral 310C in FIG. 21C is an auxiliary line for explanation.

The range of the white portion 310B is desired to be as small as possible in order to increase the recording density, but it is required to be large in order to perform the marker detection processing easily and at high speed. Therefore, the black pattern 310 when the rotation is 45 °
A range 310C is a portion 310B for sufficiently discriminating A.
It is set to enter inside.

The image magnification of the imaging optical system 200 in FIGS. 15 and 20D is the size of the data dot 316 in the data area 314, as shown in FIG. Under the conditions described below, an image is formed on 1.5 pixels. The pixel here means one pixel of the image pickup element of the image pickup unit 204. That is, one dot recorded on the sheet 182, for example, a dot of 30 to 40 μm,
It is assumed that an image is formed on 1.5 pixels of the image pickup device, which is usually 7 μm or 10 μm, through an image forming system lens. In the sampling theorem, the pixel pitch may be set to be equal to or smaller than the dot pitch, but here, for safety, 1.5 pixels are set. For an example of binarization instead of the A / D conversion described above, see 2
It is a pixel.

Adopting the above-described two-dimensional block division method has the following advantages. That is,
If the dot pitch for each dot is less than or equal to the resolution of the image sensor, the code (set of unit data blocks) can be read even if the data dot size is different; even if the image capturing unit 204 tilts with respect to the code. Readable;
It can be reproduced even if there is local expansion and contraction of the sheet, and it can be read even if rotated; unit blocks can be freely expanded in two dimensions according to the total data amount,
As a result, the code size can be freely changed;
Since each block address is added, it can be played even if you start reading from the middle of the code; if it is a block unit, you can freely change the shape of the code according to other information on the paper, such as letters, pictures, graphs, etc. Can be laid out in
Although a rectangular dot code is shown in FIG. 21 (A), it can be made into a key shape or can be deformed a little more; for example, a predetermined start code and a stop as in a bar code. No code is needed and no clock code is needed.

Further, by utilizing these characteristics, it is possible to reproduce even if there is a camera shake. Therefore, it is very easy to deal with the handy reproducing device.

That is, as will be described in detail later, on the playback device side,
Since four adjacent markers are detected and the markers are equally divided by the number of dots, normalization is performed. Therefore, there is an advantage that they are strong against enlargement, reduction, deformation and the like, and are strong against camera shake.

Regarding the dots 316 in the data area 314, for example, one dot has a size of several tens of μm. This can be up to several μm level depending on the application, but generally 40 μm.
Or 20 μm or 80 μm. The data area 314 has a size of 64 × 64 dots, for example.
These can be freely expanded or contracted to the extent that the error due to the equal division can be absorbed. Also,
The marker 310 has not only a function as a synchronization signal but also a function as a position index. The marker 310 has a size that does not exist in the modulated data, and in the case of the present embodiment, it has a circular shape and has a diameter of, for example, 7 dots or more with respect to the dots of the data area 314, or a circle having a diameter of about 7 × 7 dots. The black marker 310A is used.

Here, the tilt and rotation during reproduction will be described.

The inclination of the image pickup unit 204 means that the sheet 182 on which the dot code is printed by the reproducing apparatus.
On the other hand, although it should originally be vertically opposed, it means a state in which the user holds the reproducing device at an angle and thus becomes oblique to the sheet 182. The rotation means a state in which the image pickup area (see (A) of FIG. 4) is not parallel to the dot code written on the sheet 182.

When the above-mentioned inclination occurs, the image obtained by the image pickup unit 204 is reduced as compared with the image when vertically opposed. For example, when a tilt of 30 degrees occurs, the apparent projected image is reduced to 86.5%. That is, for example, when the block 304 has a square shape and is inclined 30 degrees in the horizontal direction with respect to the vertical direction, even if the vertical direction is 1: 1, the horizontal portion becomes 0.865 times, and the obtained block image is It becomes a rectangle. If there is such an inclination, if the clock has the original internal synchronization clock, each part operates with the evenly-spaced clock, and the resulting data may not match the original data. .

Regarding the rotation,
If we take the image of vertical, the true data rises diagonally upwards or downwards diagonally, so the true information cannot be obtained. Furthermore, when the combined state of tilt and rotation occurs, the imaging result of the square block becomes a diamond,
The condition that the horizontal and vertical data arrays are orthogonal is also not satisfied.

The marker detector 216 for solving these problems will be described below. Marker detection unit 216
As shown in FIG. 22, a marker determination unit 318 that extracts a marker from the code and determines the marker, a marker area detection unit 320 that detects an area where the marker exists,
It comprises an approximate center detection unit 322 which detects the approximate center.

The marker determination unit 318 searches for a continuous black pixel of 7 or more and 13 or less, and recognizes a case where the continuous black pixel continues for 7 rows as a circular black marker 310A.
As shown in FIG. 3, first, the image data read from the image memory 214 is binarized to identify black and white for each pixel (step S32). Then, black pixels continuous in the X-axis direction are detected on the image memory 214 (step S34). That is,
A continuous black pixel having 7 or more continuous pixels and 13 or less continuous black pixels is detected. Next, it is checked whether or not the point shifted by one pixel in the Y-axis direction from the middle pixel of the continuous first black pixel and the last pixel is black (step S36). If it continues seven times in the Y-axis direction (step S38), it is determined as the circular black marker 310A (step S40). If it is not detected in step S34 or is not a black pixel in step S36, it is not determined as a marker (step S42).

That is, it is assumed that the marker is checked on the image memory and, for example, there is a line in which seven black pixels continue. Then, it is checked whether the point shifted by 1 pixel in the Y-axis direction from the center of the first black pixel and the last black pixel is black, and if it is black, the left and right It is checked whether 7 to 13 consecutive pixels are black, and in the same manner, the pixels are shifted one by one in the Y-axis direction, and finally, in the Y-axis direction.
If it continues, the circular black marker 310A is determined.

The minimum value of 7 for checking continuous black in the X-axis and Y-axis directions is determined by distinguishing the black portion of the marker 310 (circular black marker 310A) from the modulated data. This is a lower limit value set so that the data area 314 portion and the circular black marker 310A can be distinguished from each other even if the paper is shrunk or contracted due to inclination. The maximum value of 13 is an upper limit value that is set in consideration of paper elongation, ink bleeding, and the like. This prevents noise such as dust and scratches larger than the marker from being erroneously detected as the marker.

Further, since the marker pattern 30A has a circular shape, it is not necessary to consider the rotation, so that the difference between the lower limit value and the upper limit value can be minimized and the erroneous detection of the marker can be reduced. You can

The marker area detecting section 320 determines that the range of the circular black marker 310A judged by the marker judging section 318 is
Some expansion and contraction due to tilt and changes in image magnification of the image,
Since it is deformed, it is for detecting in which area the black range is included.

In this marker area detecting section 320, the
As shown in FIG. 4, first, the temporary center pixel of the circular black marker 310A determined by the marker determination unit 318 is detected (step S52). That is, one pixel near the center of the range determined by the marker determination unit 318 is set as the temporary center pixel.

Then, it is checked that it is black in the upward direction (minus direction on the Y-axis) from the tentative center pixel, and when it becomes white, several pixels on the left and right are checked. If it is black, the upward direction is the same as above. The check is performed to the Y address where black does not exist, and the Y address is set in the Ymin register (see (A) of FIG. 25) (step S5).
4). Similarly, it is checked that it is black in the downward direction (plus direction on the Y-axis) from the provisional center pixel, and when it becomes white, several pixels on the left and right are checked, and when it is black, the downward direction is checked as above. A check is performed up to the Y address where no black exists, and the Y address is set in the Ymax register (step S56).

Next, from the provisional center pixel, it is checked that it is black in the left direction (minus direction on the X axis), and when it becomes white, it is checked that several pixels at the top and bottom are black, and it becomes black. If there is, check the left direction in the same manner as above, check up to the X address where black does not exist, and set the X address to Xmi.
It is set in the n register (step S58). Similarly,
It is checked that it is black in the right direction (plus direction on the X-axis) from the temporary center pixel, and if it becomes white, the upper and lower pixels are checked. Check up to the X address that does not exist and set the X address in the Xmax register (step S6).
0).

Xmin, Xmax, Y thus obtained
A marker area 324 is selected from the values of the min and Ymax registers as shown in the table of FIG. 25 (B) (step S62). That is, it is not the range of a square including the circular black marker 310A, but the end is removed (A) in the figure.
The area hatched with diagonal lines in FIG. The marker area 324 may be a square, but in reality, there is data around the white portion 310B of the marker 310, and the data has information such as a black data portion inside the white portion 310B due to the influence of the spatial filter or the like. It is conceivable that it will enter and enter this marker area 324 for calculating the approximate center. In order to avoid it as much as possible, it is desirable to make the marker area 324 as small and necessary as possible. In this case,
The circular black marker 310A has the same shape as that of the circular black marker 310A, that is, a circular area larger than the circular black marker 310A can be set, but in the present embodiment, the circular black marker 310A has a diameter of 7 mm.
Since it is a small circle composed of dots, it becomes a marker area 324 as shown in FIG.

The approximate center detecting unit 322 is for finding the approximate center of the black circle of the marker in the marker area thus detected by the marker area detecting unit 320. Generally, in printing or the like, there is a phenomenon that a dot expands to a target size (this is called a dot gain) or becomes small (this is called a dot reduction) due to bulging of ink. In addition, it is assumed that the ink oozes and spreads around, or that the ink bleeds to one side. In order to deal with such dot gain, dot reduction, or ink stain, the approximate center detection unit 322 obtains the center, the so-called center of gravity, in the image of the circular black marker 310A, and sets it as the approximate center. I do. Here, it is a process for obtaining the center with accuracy smaller than one pixel pitch.

First, with respect to the marker area 324 on the image, the center line on the X-axis and the center line on the Y-axis are respectively divided into the X-axis direction and the Y-axis direction of the image memory 214. By searching, find the final or approximate center. 25C and 25D are diagrams showing values obtained by accumulating each pixel in FIG. 25A and each pixel in the vertical and horizontal directions. The center of gravity is at the half of the total cumulative value, that is, the part where the cumulative values of the top, bottom, left and right become equal.

First, in the case of (C) in the same figure, for example, the result Sxl of each cumulative addition of the hatched parts in the figure is 1/2 of the total area S. If it is not satisfied and the area of the next Sxc is added to it and the area exceeds 1/2, the column S
It can be determined that xc includes the center line X including the approximate center. That is, for the X-address at the approximate center, the cumulative value of each column (Xk) is accumulated from the left side (Xmin direction), and X '+
When more than ½ of the total accumulated value at the time of accumulating one row, there is an approximate center between the X ′ row and the X ′ + 1 row. X '
When the column of X '+ 1 is divided into left and right so as to be ½ of the total area S by adding to the cumulative value up to, the approximate center is included on the dividing line.

Therefore, the ratio of the portion excluding the portion accumulated from the area of 1/2 to the X column, that is, (1/2) S-Sxl, and the accumulated value Sxc of the middle column is Δx (approximately center = X '
+ Δx).

This will be described with reference to the flowchart of FIG.

First, normalization is performed (step S72).
That is, the white data portion is set to 0 so that the addition does not affect the accumulation even if the periphery is added to each data in the marker area 324.
Then, the black data is temporarily set to 1, and the data in the image memory 214 is normalized as the data having the gradation of the multivalued data. This is for the purpose of accurately recognizing the state and accurately and accurately detecting the center of gravity, because the periphery becomes blurred by a spatial filter or the like. Next, each column Xk (k
= Min, min + 1, ..., Max), the cumulative value Sk is obtained (step S74), and the center of gravity calculation subroutine is called (step S76).

In the center of gravity calculation subroutine, (B) in FIG.
As shown in, the total area S is obtained, and 1/2 of that is Sh
And again Sl is set to 0 (step S92), i = mi
n, that is, set from the leftmost column (step S94),
It is obtained by calculating Sl '= Sl + Si (step S96). Since Sl = 0 at the beginning, here is Si
As such, Sl ′ = Smin. Next, Sl ′ is compared with Sh, that is, with a size of 1/2 of the entire area (step S98), and when Sl ′ does not exceed Sh, i is incremented (step S100), S
By setting l'to Sl (step S102) and repeating from step S96, the next column is accumulated. Then, when the cumulative result exceeds half of the total area, by subtracting Sl from S / 2 and dividing by Si, Δ
x is obtained (step S104), and i, that is, X ′ is Δ.
The sum of x is set as C (step S106), and the process returns to the upper routine.

In the upper routine, the value of C is approximately centered on X.
The coordinates are set (step S78).

Thereafter, in steps S80 to S84, the same processing is performed in each row direction to obtain the Y coordinate, and X and Y are set as the approximate centers of the markers (step S86).

The configuration for realizing such processing is as follows.
As shown in FIG. 27.

The normalization circuit 326 normalizes the white data as 0 and the black data as 1. The output of the normalization circuit 326 is accumulated by the accumulating unit 328 to calculate the total area S, is halved by the ½ multiplying unit 330, and is latched by the latch circuit 332.

On the other hand, the output of the normalizing circuit 326 is delayed by the delay circuits 334 and 336 for the blocks in the X-axis direction, the accumulating unit 338 accumulates each column in the order from the left, and the accumulating unit 340 also. Accumulation is performed for each column.
At the time of outputting the result, the portion of the central column Sxc is output.

The comparator 342 compares the 1/2 area latched by the latch circuit 332 with the cumulative value of each column accumulated by the accumulator 338. The latch 344 is for storing the timing of making a determination and the accumulation of columns up to that point. The X address calculation unit 346 causes the comparator 342 to set 1
When it is determined that the area exceeds 1/2, the latch circuit 33
2 of the area latched by 2 and Sxl latched by the latch 344, the accumulated value Sxc from the accumulator 340, and the X ′ supplied from the address controller 220 via the delay circuit 348. From the address corresponding to
The final X-address of the approximate center of the marker is calculated.

Similarly, the delay circuits 350 and 352, the accumulators 354 and 356, the comparator 358, the latch 360, and Y.
The address calculation unit 362 is used to calculate the Y address of the approximate center of the marker. In this case, the delay circuits 350 and 3
52 is composed of a line memory.

Delay circuits 334, 336, and 35 here
0 and 352 are S / 2, Sxl, Sxc, Syl, and Sy.
It is a circuit for adjusting each output timing of c to the timing required by the X address calculation unit 346 and the Y address calculation unit 362.

Next, the data array direction detection unit 218 will be described, but for convenience of description, the detailed arrangement of the dot code blocks 304 will be described first. The dot code blocks 304 are arranged as shown in FIG. 21 (B), but more specifically, FIG. 28 (A).
As shown in. That is, the block address 30
6 is divided into an upper address code 306A and a lower address code 306B, and the error detection code 312 is also divided into an upper address CRC code 312A and a lower address CRC code 312B. And the marker 31
A low-order address code 306B is arranged beside 0, and a high-order address code 306A is arranged beside it in a size larger than the low-order address code 306B. Next, a CRC code 312A having the same size as the upper address code 306A and corresponding to the higher address is added, and then a CRC code 312B having the same size as the lower address code 306B and having the lower address is added.

Also under the marker 310, the block address and the error detection data are arranged in the above order toward the lower marker.

Here, the upper address code 306A and the upper address CRC code 312A are combined to be step1.
Together with the lower address code 306B and the lower address CRC code 312B are referred to as step 2 code.

When the lower address code 306B is disassembled, on the right side of the marker 310, the data above and below each dot for indicating the lower address data (marker 3
The code that is inverted with respect to the data is written on both sides (left and right in the case of 10 lower side). Furthermore, the data margin part 36 for distinguishing from the upper and lower data areas 314.
4 are provided. The data margin section 364 may be omitted. Further, the inversion code is added not only to the lower address but also to the upper address code. Here, the dots are indicated by circles for the sake of easy understanding of the data, but actually white circles indicate that there are no dots to be printed. That is, it is not to print the white circles. Hereinafter, white circles shown in the drawings indicate the same thing.

Here, if the upper address and the lower address are all composed of 12 bits, the first 4 bits are assigned to the upper address and the next 8 bits are assigned to the lower address. It's like hitting. The data length can be appropriately changed according to the device. Basically, for all the block addresses, the order from the beginning to the upper address, and from the last address to the lower address are divided.

By providing the address code horizontally and vertically as described above, there is an advantage that even if the address cannot be detected by the address code in one direction, it can be detected by the other address code.

FIG. 29 shows another dot code arrangement.
This will be described with reference to (A). This figure omits the address code in the vertical direction of FIG. Since the address code is only in one direction, the data area can be increased and the processing speed can be increased. Since the address code is in one direction, if the address code cannot be detected, the address of the block becomes unknown, but the address can be caught by the address interpolation processing described later.

In FIG. 29A, the block address code exists only between the markers in the horizontal direction, but a dot code having the block address code only in the vertical direction may be used.

Alternatively, as shown in FIG. 28B,
The upper address code 306A may be arranged between the lower address code 306B and the upper address CRC code 312A may be arranged between the lower address CRC code 312B.

The processing will be described below with reference to the dot code shown in FIG. Only in the case of the processing peculiar to the dot code of FIG. 29A, supplementary explanation will be added.

30 and 31 are a block diagram of the data array direction detection unit 218 of FIG. 20D and a flowchart showing its operation.

The data array direction detecting section 218 receives the data of the approximate center of the marker from the approximate center detecting section 322 of the marker detecting section 216, and the adjacent marker selecting section 366 selects the adjacent marker. That is, the approximate center detection unit 32 has already been
The address of the center of each marker is mapped on one screen by the processing of 2, and the representative marker to be processed now, that is, the marker of interest is set for that (step S112), and which marker is set for that representative marker. Adjacent markers are selected to detect whether the approximate center of is closest (step S114).

The adjacent marker selection process is shown in FIG.
As shown in, the distance d between the representative marker and the adjacent marker is calculated, and the adjacent marker within the range of d ≦ dmax is designated (step S142). However, in this case, dmax is the length of the long side of the data block + α (α is determined by the expansion and contraction of the paper). Then, the approximate center addresses are sent to the step1 sample address generation circuit 368 in the ascending order of the distance d from the designated adjacent markers (step S144).
For example, in FIG. 32B, the approximate center address at the distance D2 from the representative marker is the closest, and then the distance D is next.
Since the order is the approximate center addresses of 1 and D4, and D3 and D5, the approximate center address at the closest distance D2 is sent first. If the distances d are the same, the marker is searched in the clockwise direction from the distance calculation start address, and the direction is detected in the order in which they appear (step S146). That is, D1,
The approximate center address at the distance of D4, D3, D5 is s
It is sent to the step1 sample address generation circuit 368 to perform direction detection described later.

That is, the step1 sample address generation circuit 368 generates a step1 sample address centering on the approximate center of the representative marker and the selected adjacent marker (step S116), and generates a scanning line connecting the step1 sample addresses. (Step S118), a read address is generated so that the data in the image memory 214 is sampled at points equally divided on the scanning line (step S1).
20). The address control unit 220 gives the address of this sample point as a read address to the image memory 214 to read the data.

In the above description, the data at the sampling points are approximated and output (from the image memory).
As shown in (A), when it is determined that the sample point is between the data in the image memory, the data may be interpolated from the data of the surrounding four pixels.

After the data thus read, that is, the upper address code is error-detected by the error detection circuit 370, the upper block address calculation / center calculation circuit 372 is executed.
Given to. The upper block address calculation / center calculation circuit 372 causes the next adjacent marker selection process to be performed if there is an error as a result of the error detection by the error detection circuit 370, and when a marker in two directions is detected. Since it is no longer necessary to detect the adjacent marker, the address calculation result is sent to the adjacent marker selection unit 366 to end the adjacent marker selection processing.

When the dot code shown in FIG. 29A is used, the marker selection process is terminated when the unidirectional upper address code is detected.

If the address calculation result indicates that there is an address error (step S
122), it is determined whether or not the scanning of all the sample points is completed (step S124), and if not yet, the process proceeds to the above step S118, and if the scanning of all the sample points is completed, it is confirmed whether or not there is an unsearched adjacent marker ( Step S126),
If so, the process proceeds to step S114. If not, the same process is performed for all markers. After the processing is completed for all markers, the process proceeds to marker / address interpolation processing (step S128).

The error detection circuit 370 is described in the Television Society Journal Vol. 44, No. 11, P.I. 1549 ~
P. A general one such as error detection based on a cyclic code as disclosed in “Handling Code Theory” of 1555 or the like may be used.

On the other hand, if there is no address error in the step S122, it is judged whether or not the scanning of all the sample points is completed (step S130). If not, the process proceeds to the step S118 to scan all the sample points. If it is finished, the upper address is confirmed (step S132),
The center address of step1 is calculated (step S13
4) and determine (step S136).

That is, the direction is detected by the marker closest to the representative marker (in FIG. 32B, the approximate center address is at the distance D2). The detection method determines which direction the peripheral marker is located by whether an address recorded in a dot code (step 1 code) larger than a data dot for direction detection can be recognized. The step 1 code records the upper block address and its CRC code, and is recognized if there is no error when scanning the code.

When the direction is detected, the inclination of the data block can be predicted. The step1 code has directionality, and the block address is normally recognized only when scanning from the representative marker toward the peripheral markers. Therefore, if no recognition error occurs, the block address codes in two directions are always detected. The process is repeated until the block address code in two directions is detected. Further, the data array can be estimated from the positional relationship in the two directions (see (C) of FIG. 32).

In the case of the dot code shown in FIG. 29A, the address code is detected only in one direction. At that time, the data area can be recognized from the detected line and the scanning direction (see FIG. 29B).

In the actual operation, the direction is detected from the distance D2 which is the shortest distance from the representative marker, and if the address is not recognized, the clockwise search is performed.
The same operation is repeated at the next shortest distance D1. The detection is
If it is performed clockwise, the detection continues at the distances D4, D3 and D5. Processing is performed until two directions are detected.

In the case of FIG. 29A, processing is performed until one direction is detected.

If one direction can be detected, the other direction may be predicted. For example, if D4 and D5 are in the forward direction, D2 does not exist, and the search is started from D4, when the address is confirmed in D4, it is predicted that the address can be recognized in either D3 or D5.

The direction detecting process as described above will be described in more detail with reference to FIG.

The approximate center detector 322 of the marker detector 216
The approximate center of the representative marker detected in step S1 is defined as a dot A5 on the upper left side of the figure, and eight sample points A1 to 1.5 dots (this can be changed appropriately depending on the processing) from the dot A5.
A4, A6 to A9 are generated by the step1 sample address generation circuit 368. Similarly, a marker for which direction detection is to be performed, for example, the approximate center of the distance D2 (dot B on the upper right side of FIG.
A sample address is generated around 5).

Here, the reason why the interval is set to 1.5 dots will be described.

In the process of obtaining the approximate center of the marker, the description has been made so that the difference from the center is within 1 dot, but it is assumed that a problem such as ink bleeding does not occur. Considering ink bleeding etc., the detection range is ±
It is set to 1.5 dots.

The address control unit 220 draws a certain line between the addresses of both markers. First, a scanning line is drawn on the dots A1 and B1. And by providing a sample clock so that the upper address can be sampled,
Data sampling of the image memory 214 is performed.

As shown in FIG. 28A, since the CRC code 312A is added next to the upper address code 306A, if the data sample can be read correctly, the upper address On the other hand, the error detection result in the error detection circuit 370 is detected in the form that there is no problem, and if it cannot be read correctly, it is determined that there is an error.

Then, similarly, the dots A1 and B2,
Scanning lines are sequentially drawn in the order of A1 and B3, A1 and B4, and it is checked whether or not an error has been detected. Since there are a total of 9 positions on the representative marker side and 9 positions on the detected marker side, 81 processes are performed in total.

When an error occurs in all the 81 processes, it is determined that there is no direction code in that direction, that is, the detection side marker is a marker other than the array (a marker that has been erroneously detected).

For example, in FIG. 33A, the dot A1
If data is taken at each sample point on the scanning line (indicated by a dotted line) drawn for B7 and B7, the sample point indicated by a broken line circle in the figure is out of the data, resulting in erroneous detection. Especially, as described above, since the inverted codes are provided on the upper and lower sides of the address data dot, an error always occurs.

On the other hand, when the dots A5 and B5 are connected,
Since the data is properly collected, there are no detection errors,
Therefore, it is recognized that there is a code in this direction.

Although inverted codes are provided above and below to facilitate error detection, this need not necessarily be provided above and below. For example, white codes are written above and below the address data dots, and address data dots are written. Can be in a format such that black data continues for only the last few dots. In this case, the end on the detection marker side is always black data and the outside is a white margin, so that the data error can be correctly detected. Also, in the case of the inversion code, it is not necessary to provide it in the entire inversion code portion, and it may be provided in a part of both sides ((C) of FIG. 28).

Here, the size of dots will be described. As shown in (B) of FIG. 33, the size of each dot of the high-order address code 306A is n dots, and step1
If the code width is m dots, the relationship between m and n is s
At the inner edge of the tepl sample address, a width of 2 dots is provided with respect to the center and a diagonal line is drawn, and the width m is determined by how much the upper address code 306A is provided.
Is the long side and the diagonal is the diagonal, and the height of the rectangle is n. That is, if m is determined, n is inevitably determined. Even if all the space between the inner ends of the step1 sample addresses is set to 2 dots, n dots have a width of up to 2 dots. Although the horizontal width of one dot is not determined, a horizontal width that facilitates data recognition is preferable.

Note that the above-mentioned two dots have a range in which, for example, the scanning line connecting the dots A5 and B5 hits, but the line connecting the dots A6 and B4 and the line connecting A2 and B8 do not hit. Stipulates that you can. If it is larger than that, for example, dot A
In the case of hits with 5 and B5, the hit may occur even if A2 and B8 are subtracted, and the detection as the center spreads. This value can also be changed according to the device.

In the example of FIG. 33A, the dot A
5 and B5 are hit, but dot A4 is the same
If, for example, a line connecting B4 is hit, then at the next step 2 of center detection, the center of dots A4 and A5 is used as a starting point and the same search is performed with that as the center. Processing will be carried out.

Another method is also conceivable. This will be described with reference to FIG. Here, A4 and A5, one side is also B4 and B5
When is hit, the sample address (A4
1-A45, A51-A55, B41-B45, B51
B55) may be used as the sample address of the next step2. In this case, since the number of sample address points in step 2 is increased from 9 to 10, the number of processes is also 81 to 1.
00 (the number of scanning lines). But A4 and A5
Since the process of deriving the midpoint and the predetermined sample points are used, 9 step2 points centering on the midpoint
The process of generating the sample address of is eliminated. Overall, the processing seems to be reduced.

Further, assuming that the exact center of step2 is between A4 and A5, A42 to A44 and A52 to
If the address detection processing is performed on the scanning lines connecting A54 and B42 to B44 and B52 to B54, the number of processings can be reduced from 81 to 36 (6 × 6).

With the above processing, the rough center at step 1 is obtained.

As described above, by detecting the CRC, it is detected whether or not the data blocks are properly arranged in that direction. FIG. 32 (B)
In this case, of course, the marker at the distance D2 becomes a erroneously detected marker, so when looking at the data direction in that direction, there is no upper address code, so there are 81 possible detections. Then, all errors will occur in that direction, and it will be determined that there is no direction.

When it is thus determined that there is no D2, the next shortest distances are D1 and D4, but since it turns clockwise with respect to the marker of interest now, processing is next performed for the distance D1. As described above, in terms of the data array, the determination can be made only from left to right and from top to bottom. In this case, therefore, the processing is performed in the direction from the representative marker to the marker at the distance D1. , From the opposite direction, ie C
Since the RC code is read first and then the address code is read, this is naturally determined to be an error.
Therefore, it is determined that the distance D1 has no direction.

Next, the distance D4 is determined. When D4 is read along the distance D4 from the representative marker, it is read in the order of the address code and the CRC code. Therefore, it is determined that D4 has directionality. That is, no error occurs.

Next, the slips to be judged are the distances D3 and D5 which are equidistant. On the other hand, since it is clockwise, the processing is first performed from the distance D3. This D3
Also, since the CRC code is read first as described above, it is detected that there is no directivity. Then, finally, the distance D5 is read, and it is determined that there is a direction here.

As a result, since the distances D4 and D5 are read, the hatched portion in FIG.
It can be recognized that the data for the block address described in the portions of the distances D4 and D5 is written. Finally, if two representative directions are detected for one representative marker, the directions of the blocks can be detected, and therefore the processing is performed until the two directions can be detected.

In the case of the dot code shown in FIG. 29A, only one direction is detected. The processing is performed until one direction (D5 in FIG. 29B) is detected.

If an error occurs by performing processing on all of the above five directions, the direction detection processing is performed on the marker in the diagonal direction. In this case,
In order to prevent an increase in the number of processes, those outside a certain range are not processed, and the address information that cannot be obtained is
Obtain necessary information by marker and block address interpolation processing.

As described above, the block address is not modulated. However, when the block address is modulated, it is naturally necessary to perform the process of demodulating after recognizing the block address code.

In the above description, it is assumed that the error detection of the upper address is used to judge whether or not there is directionality. However, for example, instead of the upper address CRC code,
Use a directional pattern such as "11100001" and adopt a method that recognizes that there is a directional marker in that direction when "11100001" is detected by pattern matching. You can also

In the above direction detection, it is not necessary to search for adjacent markers in a clockwise direction for all markers, and the next block is
An operation for recognizing the higher-order address code may be performed in that direction. That reduces the number of processes.
Further, even when an abnormality occurs in the detection of the upper address, it may be recognized that the code exists in the direction obtained by detecting the peripheral direction.

Next, the block address detection, error determination, and accurate center detection unit 300 will be described with reference to the block diagram of FIG. 35A and the flowchart of FIG.

When the upper block address calculation and center calculation circuit 372 of the data array direction detection unit 218 can detect the upper address, the upper block address is detected as the next block address, the error is determined, and the correct center is detected. It is sent to the block address calculation and center calculation circuit 374 of the unit 300. Further, since the rough center at the time of detecting the upper address is known, this center address is set to step2.
Guide to the sample address generation circuit 376 (step S1)
52).

Step 2 sample address generation circuit 37
6 generates this rough center sample address (step S154). That is, as shown in (B) of FIG. 35, the approximate center (center of direction detection) found earlier.
On the other hand, as in the above case, 8 points and sample addresses are placed outside. Then, eight points are similarly provided for the marker for which the directionality is found, a scanning line is similarly drawn (step S156), and processing is performed as to whether or not the lower address can be detected. In this case, the data interval for forming the sample address is defined every 0.5 dots in the present embodiment, but can be changed as appropriate according to the specifications of the device.

Then, the address control section 220 reads the data from the image memory 214 based on the generated sample address, and leads the data according to this sample point to the error detection circuit 378 (step S158). As in the case of direction detection (as shown in FIG. 5A), when the sample points are between the data in the image memory, 1
The data may be interpolated and derived from surrounding data instead of representing the data. If an error is found in the error determination (step S160), it is determined whether or not scanning of all sample points is completed (step S162). If not, the process proceeds to step S156, where scanning of all sample points is completed. If so, after the addresses have been detected for all blocks, the process proceeds to marker / block address interpolation processing (step S164).

On the other hand, if there is no address error in the above step S160, it is judged whether or not the scanning of all the sample points is completed (step S166). If not, the process proceeds to the above step S156 and the scanning of all the sample points is performed. If completed, the lower address is confirmed (step S168),
Determine the exact center (step 2 center) (step S
170).

That is, the error detection circuit 378 detects an error, and if an error is found in the error determination, the process goes to the next step. The block address calculation and center calculation circuit 374 is provided with a start and end address at the time of center detection, that is, a signal indicating which point is connected and which point is currently connected from the address control unit 220. Judge whether or not. When no error is detected, the block address calculation / center calculation circuit 374 combines the derived lower address with the upper address sent from the upper block address calculation / center calculation circuit 372 to obtain a block address. , To the next marker and address interpolation unit 302. Similarly, the center address is also derived to the marker / block address interpolator 302.

In FIG. 35B, the setting of 0.5 dot means that the center point finally obtained by this processing is detected by detecting the sample points in the range of 0.5 dot ( This is because the difference between the center of direction detection) and the true center falls within the 1/4 dot range. If it falls within the 1/4 dot range, if the sample points formed by the above processing are taken,
The data in the data area can be played properly.

Since the smallest dot of the step 2 code is 1 dot, and a data arrangement smaller than that is meaningless as data, it is formed by 1 dot.

As in the case of the step1 code,
Inversion codes may be provided above and below the address data dots, or black data may be provided at several dots at the end of the address data dots, and the surrounding area may be used as a margin portion. In addition, a data margin part 364 for distinguishing the address code and the data code
Since the area to be distinguished from the data area 314 is very likely to be mistaken for a marker even if it overlaps with the black area, the data margin 364 is not provided and the area is directly entered from the inversion layer. May be.

As shown in FIG. 35B, as a result, the lower address and the upper address have a data length of approximately ½ of the total data length, and CR of the same size.
C code is added. The reason is that the data length is set so that it is possible to detect a burst error in such a state that noise is all over the address length, ink is attached, or the like. The ratio of this data length can also be changed as appropriate.

By the tree search processing as described above, that is, the detection method of obtaining a rough center and then a finer center, the accurate center for sampling the data in the data area 314 and the block address are recognized. It was done. That is, by performing the process of tree search, the amount of processing is greatly reduced, and the amount of processing and the processing time are reduced, compared with the case where the sampling is performed at a fine pitch from the beginning.
Further, by using the block address for the direction detection and the accurate center detection, the redundancy of the total data amount can be reduced.

Next, the marker / address interpolation unit 302 will be described with reference to FIG. In the figure, it is assumed that the black marker portion around the block B2 is detected in response to the error that the marker for the block B2 is not detected or the address is not detected.

In this case, first, a line is drawn which connects the centers of the marker of block B1 and the marker of block B3, and a line which connects the centers of the marker of block A2 and the marker of block C2, and the intersection point is drawn. Use as a prediction center. Then, the address can be detected and processed from the predicted center point toward the marker of the block C2 and the marker of the block B3. Further, since the array is known without performing address detection, if block B2 exists below block B1, the address of block B2 is set from the surrounding addresses, so there is no need to detect it. Can also be estimated. That is, it is possible to detect the address and the center of the marker of the block, which has been predicted and cannot be predicted, from the surrounding processing.

Interpolator 302 for marker and block address
Guides the address data normally read, the address interpolated with the center position, and the information of the predicted center to the address control unit.

When the image data is captured in the image memory 214 as shown in the figure, and the scanning direction is the arrow direction, the upper left corner is set as the first representative marker and the processing is performed from that point. The center is detected sequentially in the vertical direction,
Eight (blocks A1 to
The centers of the A4 marker and the blocks B1 to B4) are obtained. Then, when the center of the next column is detected, the centers of the markers of the blocks B1 to B4 are already known, and therefore no processing is performed on them, and the markers of the blocks C1 to C4 are targeted for those centers. The rough center, the center of step1, and the center of step2 are obtained. Therefore, as described above, 81 scanning lines are not necessary, and once the center is obtained, it is sufficient to perform processing so as to sample the 9 points in the latter stage. Ie 1
The center can be obtained by eight types of processing. in this way,
Although there are many processes only at the beginning, there is an advantage that the processes after that are reduced.

In the case of the dot code shown in FIG. 29A, first, the upper left A1 is used as the representative marker A1,
Direction detection processing is performed in the horizontal direction with B1 and C1. Processing is A
Once the marker centers of 1 and B1 are obtained, the center detection processing of C1 may be nine kinds of processing. In order to determine that the block under A1 is A2, since there is no address code, processing is performed as described below.

That is, the block size may be judged from the lengths of the A1 marker and the B1 marker, and detection may be started from the marker at an appropriate position from the predicted block size, or the marker immediately below A1 may be detected. First, the processing may be performed using the representative marker. Then, the block in which the horizontal block address matches the detected block address may be set as A2. When the detection of the direction of two steps (the step of A1 and the step of A2 in the figure) is completed, the direction can be predicted in the process of the vertical direction (the process of selecting the marker of A3), so only the marker in that direction is detected. It suffices to perform processing. Even if there is a marker that is erroneously detected, it is possible to perform the processing excluding it.

Next, the address control section 220 of FIG. 20A will be described with reference to the block diagram of FIG. 37B.

First, in the address control unit 220,
The data of the A / D conversion unit 210 is stored in the image memory 214 at the address generated by the write address generation unit 380 that generates an address when writing the data from the A / D conversion unit 210 to the image memory 214. .

Then, as described above, the marker detector 2
16, it is necessary to generate an address in each of the data array direction detection unit 218, block address detection, error determination, accurate center detection unit 300, and marker / address interpolation unit 302, and the address generation unit 382 for that purpose.
~ 388 are configured. In this case, the marker detection address generator 382, the data array direction detection address generator 384, the block address detection, the error determination,
In the accurate center detection address generation section 386, the corresponding marker detection section 216 (the internal marker determination section 31
8, marker area detector 320, approximate center detector 32
2) Addresses are generated by exchanging information with the data array direction detection unit 218, block address detection, error determination, and accurate center detection unit 300. Further, the interpolation processing address generation unit 388, for a block in which four markers around the block exist, an address (hereinafter, referred to as a marker address) in which an accurate center of each marker is associated with the image memory, and the number of data. Further, the memory read address of the interpolated address coordinate data and the pixel data in the periphery of the block is equally generated.

The selection circuit 390 selects these address generators 382 to 388 at each timing and supplies them to the lens aberration distortion correction circuit 392. Then, the lens aberration distortion correction circuit 392 receives the distortion information of the lens aberration from the lens aberration distortion memory 224 and converts (corrects) the selectively supplied address,
The read address is given to the image memory 214 via the selection circuit 394.

Next, another embodiment of the marker determination unit 318 in the marker detection unit 216 will be described with reference to FIGS. 38 (A) to (C).

In the above-described embodiment, when the size of the dot code is determined, the image forming optical system 200 forms an image so that one dot corresponds to 1.5 pixels of the image pickup element of the image pickup unit 204, and the marker is used. The determination unit 318 finds a two-dimensionally continuous black pixel and determines it as a marker. On the other hand, in the present embodiment, when there are codes with different dot sizes, for example, codes with a dot size of 20 μm, 40 μm, and 80 μm, the image magnification in the imaging optical system 200 is not changed. It allows you to play each code.

That is, the paper quality, the sheet property, the ink level, and the printing level are different among various applications, and therefore, the dot size code corresponding to each application is used. For example, it is conceivable that 20 μm will be used when the recording density can be extremely increased, and 80 μm will be used depending on the application using a very rough low-cost sheet having poor quality. In such a situation, determine its size,
The purpose is to play this code correctly.

That is, as shown in (B) of the figure, there are a circular dot size 20 μm marker, a dot size 40 μm marker, and a dot size 80 μm marker.
The reproducing apparatus to which this embodiment is applied is, for example, 20 μm.
It is assumed that the device for efficiently reproducing the code of m, that is, the device having the magnification of the image forming system capable of decoding more information in one imaging. Then, in an apparatus which is imaged at an image magnification of 1.5 times with respect to a dot of 20 μm, each code of 40 μm and 80 μm can be reproduced without changing the image magnification of the image forming system. That is the purpose. However, the size of the marker shown in (B) of the figure was set to a diameter 7 times the size of each dot.

Therefore, as shown in (A) of the figure, first, the code of the maximum dot size to be selected is initialized (step S182). For example, 80 μm, 40
When there are codes of μm and 20 μm and it is desired to reproduce all of them, the maximum size is set to 80 μm. This may be set by key input by the user, or if it is determined that there are three types of 80 μm, 40 μm, and 20 μm, and it is possible to handle only that size, the device itself The size may be set to a large size of 80 μm.

Then, the provisional center is obtained by making a determination with the marker determination formula shown in FIG. 13B (step S184).

That is, if a code is made using 7 dots of each dot size as a marker, then at that time, since the imaging optical system has an image magnification of 1.5 times, the code of 20 μm is used. In case of, the diameter is 10.5 dots, and in case of 40 μm code, it is 21 dots, 80 μm
In case of the code of, it becomes 42 dots. Therefore, if 7 pixels or more and 12 pixels or less, or black pixels are two-dimensionally continuous, 20
Judged as a marker of the code of μm, 14 pixels or more 24
If the number of pixels is less than the pixel, it is judged as a marker of 40 μm code,
Those having 29 pixels or more and 47 pixels or less are determined to be markers having a code of 80 μm.

The value of this pixel is calculated by the following equation.

R = s × d × m int (r × 0.7) ≦ R ≦ int (r × 1.1 + 1) where r: number of pixels corresponding to marker diameter (= 7) s: dot size (20 μm, 40μm, 80μ
m) m: Image magnification of image forming system (= 1.5) d: Number of dots of diameter of marker R: Diameter of marker in binary image Number of pixels 0.7: Reduction rate due to inclination, dot rejection, etc. 1. 1: An enlargement ratio due to dot gain or the like.

First, at step S182, 80
Since the marker is initially set to a μm code marker, it is checked in step S184 if the marker determination formula is an 80 μm code marker, and a marker of that size (a marker formed of 80 μm dots) is determined. For those that are determined to be present, the provisional center is obtained.

Next, the number of the markers is checked, and it is checked that there are four or more markers (step S18).
6). This means that one block is surrounded by four markers, and thus it is determined whether or not there are one or more blocks.

Then, it is confirmed whether the marker has a predetermined positional relationship with the adjacent marker as shown in (C) of the figure, that is, whether the marker is aligned (step S18).
8). That is, the marker B located in the vicinity of the target marker A, the marker C located in the vicinity of a position separated by the distance D in the direction perpendicular to the direction connecting the markers A and B from the target marker A, and the marker B as a reference. The marker D located in the vicinity of the position separated by the distance D in the same direction as the direction of the markers A to C is detected.
If they exist, it is determined that the code is, for example, 80 μm in this case.

In step S186, 8
If there are not four or more markers of 0 μm code, or if it is determined in step S188 that they are not aligned, it is determined that this is not an 80 μm code, and one smaller code, 40 μm in this case
(Step S190), the process returns to step S184, and the marker is determined once again.

If it cannot be determined even in the determination of the smallest size, it means that it is not a code, or even if it is a code, it cannot be reproduced, and the process is terminated. In this case, it is preferable to proceed to a process of issuing a warning such as issuing an alarm.

Next, another embodiment of the marker determining section 318 will be described. That is, a method of determining the marker pattern and the modulated data by dilation, which is a general image process, will be described. Here, the dilation process is a process of converting a black pixel near a white pixel into a white pixel. Specifically, for example, the pixels around 3 pixels of the pixel of interest (7 × 7 pixel area centering on the pixel of interest) are checked (black and white determination), and if even one pixel is a white pixel, the pixel of interest is made a white pixel. The conversion process is performed for all pixels on the image.

First, binarization processing is performed on the data in the image memory.

Next, by the above dilation processing, only the data portion of the code image is entirely converted into white pixels, and the pattern portion of the marker is converted into an image which is smaller than the initial size by the number of dilated pixels. It

Next, the address in the image memory of the change point from the white pixel to the black pixel on the image and the number of consecutive black pixels from the pixel are counted, and the information is classified for each marker from the information. , The temporary center address and the marker existence range are detected. Then, the approximate center detection process is performed.

As a result, the marker determination and the marker existing range can be detected at high speed.

In the case of a code in which the marker center is uniformly deformed like the dot gain and dot reduction described above for the marker, the temporary center address obtained by the marker determination is set as the approximate center as it is. You can also

The process of step S184 of FIG. 38A may be the above process.

When the above-mentioned A / D conversion unit is binarized by a comparator, in the above marker determination processing,
The binarization process can be omitted.

Next, a light source integrated image sensor applicable to the detection unit 184 of the reproducing apparatus shown in FIG. 15 and FIG. 20D will be described. FIG. 39 is a diagram showing the structure thereof. For example, the light emitting cell 3 is formed beside the light receiving cell 396 by a compound semiconductor such as an LED or an electroluminescence element.
98 is formed on-chip. Between the light receiving cell 396 and the light emitting cell 398, an isolation (light shielding) portion 400 in which a cutter is actually formed on the wafer to form a groove and which is non-transmissive, for example, metal is embedded is provided. The isolation section 400 can eliminate the problem that the light emitted from the light emitting cell 398 directly enters the light receiving cell 396.

In such a structure, the light emitting cell 39
Light emission is controlled according to the light emitting cell control pulse signal as shown in the timing chart of FIG. The light receiving cell 396 sends the accumulated charges to the adjacent vertical charge transfer register 402 by applying the charge transfer gate pulse signal to the charge transfer gate (not shown). The vertical charge transfer register 402 sends the accumulated charges to the horizontal charge transfer register 404 line by line with a vertical charge transfer pulse.
The horizontal charge transfer register 404 outputs accumulated charge pixel by pixel via the buffer amplifier 406 in response to a horizontal transfer clock signal.

Next, in the circuit of the reproducing apparatus described above, the preceding stages of the demodulation circuit are implemented by analog circuits, and
An embodiment in the case of a chip will be described with reference to FIG. In the present embodiment, as the imaging unit,
For example, by using an XY address type image pickup unit 408 represented by CMD as disclosed in Japanese Patent Laid-Open No. 61-4376, a memory is unnecessary and therefore a circuit system can be reduced. It becomes possible to do. In order to address scan the XY address type image pickup unit 408, an X decoder 410 and a Y decoder 41 are provided.
2 is prepared.

In a normal XY address type image pickup unit, CC
Unlike D, one line is read out and then this line is reset to read out the next line, that is, another line enters the exposure period while reading another line. Target. However, such a reading method has a demerit that an extra portion is exposed when external light enters during the image pickup time. Therefore, in the present embodiment, the XY address method is used, and In combination with the element shutter, the operation is performed such that exposure is performed only when external light comes in, that is, when exposure is to be performed, and other areas are not exposed.

The image sensor scan address generation and element shutter control unit 414 thus generates an element shutter pulse for providing an element shutter-like operation in the XY address system, and a reset pulse for resetting all pixels. .

The X-decoder 410 and the Y-decoder 412 are circuits for turning on any one of the elements in response to the X-address and the Y-address from the image-capturing element scanning address generation and element shutter control section 414. Normally, it is composed of a shift register or the like, but in the present embodiment, it is a type of selector in which any one element can be turned on by a signal from the image sensor scanning address generation and the element shutter control unit 414.

The reset pulse in this embodiment corresponds to the image sensor reset pulse in the timing chart of FIG. 16, and the image sensor is reset before the exposure, and the reset pulse is reset during this reset period. By going high, switch 416 is switched, drawing all charge towards negative power supply 418.

The element shutter pulse is generated in such a manner that it can be gated from after the end of the reset pulse to after the end of exposure, as shown by the broken line waveform in FIG.

Reading is performed in the same manner as a normal pulse.
The respective elements are sequentially turned on, and the signal charges are transferred to the current / voltage conversion amplifier 42 via the reset selection switch 416.
After being amplified by 0, it is supplied to the marker detection unit 422. The marker detection unit 422 is the same as that described above, and the data detected by the marker is stored in the register 424. The θ detection unit 426 calculates the inclination based on the contents of the register 424 as in the direction detection unit described above. For example, in the circuit shown in FIG. 20D, the θ detector 426 corresponds to the data array direction detector 218, and the next data interval controller 428 and the image sensor scan address generator / element shutter controller 414 address the data. It corresponds to the control unit 220.

Then, the coefficient for interpolation generated from the coefficient generating section 430 under the control of the data interval control section 428 is multiplied by the charge read by the multiplication circuit 432.
All are added in the adder circuit 434. That is, the output of the adder circuit 434 is sampled and held by the sample and hold (S & H) circuit 436, and the switch 43
It is returned to the addition circuit 434 via 8. This behavior is
When the data is sampled after the direction and the scanning line are determined, it is performed to perform data interpolation as shown in FIG. In FIG. 5A, D6, D7, D10, and D11 are interpolated by applying coefficients to obtain Q data. The value thus interpolated is further added to the S & H circuit 440.
The sampled and held value is sampled and held, and the sampled and held value is binarized by the comparator 442 and the threshold value judgment circuit 444.

Each image pickup device (pixel) of the XY address type image pickup unit 408 will be described in more detail. As shown in FIG. 41A, each pixel is composed of two CMD elements, and the element shutter pulse has the first CMD element 4
Capacitor 4 that enters 46 and is stored for the element shutter
The charge is accumulated at 48. Then the second CMD
The Y-decoder 412 drives a pulse for reading the element 450 to select a line, and the horizontal selection switch 452 reads the electric charge for each pixel.

At the time of exposure, the element shutter pulse causes the first CMD element 446 to operate as an element shutter, and charges are accumulated in the element shutter capacitor 448. When the charges are accumulated in this way, the charges are shielded, a line for reading is applied from the Y decoder 412 to select a line, and the second CMD element 450 is turned on by the horizontal selection switch 452 to read one pixel at a time.

When resetting the electric charges, all the horizontal selection switches 454 are turned on by the reset pulse output from the image pickup device scanning address generation and element shutter control unit 414, and the reset selection switch 416 is set to the negative power source 4.
Set to 18 side. Since the source of the CMD element 450 has a negative voltage, the charges accumulated in the element shutter capacitor 448 and the gate of the CMD element 446 move to the negative power source and are reset.

In addition to the above operation, the voltage of the element shutter pulse and the read pulse can be reset by applying a slightly higher voltage at the same time.

In the case of an ordinary image pickup device, dark current is a problem, but in the case of the present embodiment, it is shown in FIG.
Since the element shutter pulse shown in (1) is exposed only during the high period and the charge is read out immediately, the dark current accumulation time is actually very short, and therefore S / Compared to the operation of other image pickup devices, the N ratio is advantageous. In exposure, a sufficient amount of light is given even in this short exposure period, so that the signal level remains the same and the S / N level with respect to the dark current becomes small. Therefore, by applying this embodiment, It is possible to set a considerably large gain for the output degree of the current-voltage conversion amplifier 420 in the subsequent stage.

In the present embodiment, the pixel structure for performing the element shutter operation as described above is used.
It is also possible to use a CMD element capable of element shutter operation as disclosed in Japanese Patent Publication No.

Next, referring to FIG. 42, the above X
A circuit using the Y address type image pickup unit 408 is used as a three-dimensional I
An embodiment constructed in C will be described. It should be noted that the present embodiment is a case of an audio information reproducing apparatus.

[0285] This is a CM for the surface of the sheet 182.
The imaging unit layer 454 including the D 408, the X decoder 410, and the Y decoder 412, the detection unit layer 456 that detects data formed by stacking on the imaging unit layer 454, and the stacking on the detection unit layer 456. Then, the output processing layer 458 is formed. The output processing layer 458 includes the demodulation unit 19
0, an error correction unit 194, a decompression processing unit 256, a data interpolation circuit 258, a D / A conversion unit, an output buffer 266, and the like, and the decoded audio information is reproduced as sound by an audio output device 268 such as an earphone.

Of course, the output processing layer 458 can also be configured to reproduce the multimedia information including the image information, as described above.

By using a three-dimensional IC in this way, processing up to sound output can be performed with one chip.
The circuit scale becomes very small, and it also leads to cost reduction.

Next, various configuration examples of the pen-type multimedia information reproducing apparatus will be described.

For example, the pen-type information reproducing apparatus can be provided with a switch for instructing the timing of loading the dot code.

FIG. 41 (B) is a diagram showing an example thereof. This pen type information reproducing apparatus has a light source 198 and an image forming device in the reproducing apparatus shown in FIG. 15 or 20 (D). A detection unit 184 including the optical system 200, the spatial filter 202, the imaging unit 204, the preamplifier 206, and the imaging unit control unit 212.
Is provided at the tip thereof, and the scan conversion unit 186, the binarization processing unit 188, the demodulation unit 190, the data error correction unit 194,
The expansion processing unit 256, the data interpolation circuit 258, etc.
The image processing unit 460, the data processing unit 462, and the data output unit 464 are incorporated. Then, the audio output device 26
I have 8 earphones. In this figure,
Although only an output device for audio information is shown, if a processing unit for images, characters, line drawings, etc. is built-in, it is of course possible to connect an output device corresponding to it (see the following pen-type information reproducing device). The same applies in the explanation).

A touch sensor 466 is provided on the side surface of the pen-type information reproducing device. As the touch sensor 466, for example, a piezoelectric switch, a micro switch, a piezoelectric rubber or the like can be used, and a switch having a small thickness of 0.6 mm or less is known. The control unit as the image capturing unit control unit 212 starts capturing the dot code as described above in response to the touch sensor 466 being pressed by the finger. And
When the finger is released from the touch sensor 466, the capturing ends. That is, the touch sensor 466 is used to control the start and end of the dot code acquisition.

Reference numeral 468 in the figure is a battery as an operating power source of each unit in the pen-type information reproducing apparatus.

The touch sensor 466 is not limited to being pressed by a finger, but the same function can be achieved even if it is attached to the tip of the pen-type information reproducing apparatus as shown in FIG. .

That is, when the user places this pen-type information reproducing device on the sheet 182 to manually scan the dot code printed on the sheet 182, the touch sensor 46
6 is turned on, the control unit 212 recognizes it and starts reading the dot code.

In this case, since the tip of the pen-type information reproducing device moves in contact with the sheet surface during scanning, in this example, the tip of the touch sensor 466, that is, the surface in contact with the sheet surface, is made of smooth resin or the like. Is preferably coated to provide smooth movement during manual scanning (movement).

Further, the detecting unit of the pen type information reproducing apparatus may be further provided with a mechanism for removing specular reflection.

FIG. 44A is a diagram showing the structure thereof. The first polarization filter (polarization filter 1) 470 is arranged on the front surface of the light source (light source such as LED) 198, that is, on the irradiation side. A second polarization filter (polarization filter 2) 472 is arranged on the front surface of the imaging optical system (lens) 200.

For example, the first polarization filter 470 has a polarization filter film 474 as shown in FIG.
Is formed by cutting out into a donut shape, and the second polarization filter 472 has a different polarization filter film 47.
6 can also be used, and for example, as shown in (C) of the same drawing, an inner portion of the polarization filter film 474, which is obtained by cutting out the first polarization filter 470, can be used.

Then, the first and second layers thus formed
The polarization filters 470 and 472 are arranged such that the pattern surface (polarization plane) of the second polarization filter 472 is orthogonal to the pattern surface (polarization direction) of the first polarization filter 470.

As a result, the random light emitted from the illumination light source 198 has its plane of polarization limited by the first polarization filter 470 and is irradiated with, for example, P waves. The polarization plane of the specular reflection component is preserved as it is and returned from the sheet surface as a P wave. However, since the polarization plane of the second polarization filter 472 is orthogonal to that of the first polarization filter 470, The reflection component is blocked by the second polarization filter 472. On the other hand, in the case where the light emitted from the first polarization filter 470 returns to the actual dots, that is, the sheet surface and returns as the brightness information of the paper surface, the polarization plane becomes random. Therefore, like this, once you enter the page, the black and white information,
Alternatively, the signal returned as color information has both P and S components. Among them, the P component is similarly cut by the second polarization filter 472, but the S component orthogonal to the P component passes through the second polarization filter 472 and actually passes through the lens 200. An image is formed on the imaging unit 204. That is, the reflected light from which the regular reflection component is removed is guided to the image pickup unit 204.

In this case, a 1/4 λ plate 1230 is arranged on the front surface of the spatial filter 202, and the image light that is once incident as linearly polarized light is converted into circularly polarized light and is input to the spatial filter 202. This is because the spatial filter normally utilizes the birefringence of quartz, so for linearly polarized light,
This is because that effect cannot be obtained. In this example,
The quarter-wave plate 1230 is arranged on the front surface of the spatial filter 202, but is not limited to this.
It may be arranged at any place between the polarization filter 472 and the spatial filter 202 which is easy to install.

As a structure for removing the specular reflection component in this way, a structure as shown in FIG. 45 can be considered. This is because the first polarization filter 470 is connected to the light source 19 described above.
Instead of arranging in the vicinity of 8
Using the transparent resin optical waveguide material 480 provided with 78, the light from the light source 198 is guided to a position very close to the sheet surface to illuminate the sheet (dot code). It is arranged in the light emitting part of the. In this case, the first polarization filter 470 is arranged so that the light orthogonal to the second polarization filter 472 is transmitted.

Incidentally, here, the transparent resin optical waveguide material 480 is used.
Is advantageous in that the light source 198 and the outer shape can be made as thin as possible, and the incident angle can be made shallow so that the regular reflection component can be reduced.

However, since the specular reflection component still remains due to the swelling of the ink, the swelling of the sheet surface, etc., a polarization filter is provided in order to eliminate it more efficiently.

Further, instead of the second polarization filter 472, an electro-optical element shutter 1220 such as a liquid crystal shutter or a PLZT shutter may be provided. As shown in FIG. 44D, the electro-optical element shutter 1220 includes a polarizer 1221 as a polarization filter, a liquid crystal, and a PL.
It is composed of an electro-optical element 1222 such as ZT and an analyzer 1223 as a polarization filter. In this case, the shutter 1
By disposing the shutter 1220 so that the light distribution direction of the polarizer (polarization filter) 1221 of 220 is the same direction as the second polarization filter 472, the specular reflection removing effect can be obtained.

Furthermore, the shutter function allows IT-CC
There is a merit that a frame reading is possible with an image sensor compatible with field reading such as D, or that all pixels are simultaneously exposed even with an XY addressing type image sensor such as CMD.

Next, an example in which the light source 198 is made more efficient and the device is made slim will be described.

FIG. 46 (A) is a diagram showing the structure thereof, and like the example of FIG. 45 (A), the surface is covered with the mirror coat 4
An acrylic transparent resin optical waveguide material 480 having 78 is provided. As shown in FIG. 46 (B), this acrylic transparent resin optical waveguide material 480 is formed in a truncated cone shape, and a screw portion 482 is provided on the upper portion (the end portion on the widening side) thereof. It is adapted to be screwed and attached to the housing 484 of the pen-type information reproducing device. Also, this screw portion 482
The surface mirror coat 478 is not applied to the inner portion in the vicinity, and the light source 198 is provided in the portion 486. That is, the light source 198 is provided as an LED array in which LEDs are mounted on a thinly cut flexible substrate 488, and the LEDs are arranged in a ring shape, and the LED array is attached to the portion 486 without the surface mirror coat. . Then, as shown in (C) of the figure, the lower portion (tip portion) of the acrylic transparent resin optical waveguide material 480 is cut, and the portion 49 where the surface mirror coat 478 is not applied.
0 is formed. Therefore, the light from the light source 198 enters the transparent resin optical waveguide material 480 through the mirror-uncoated portion 486, is reflected by the surface mirror coat 478, passes through the optical waveguide material 480, and the surface mirror coat of the tip portion. It goes out from the blank portion 490 and is irradiated on the dot code on the sheet.

As shown in (D) of the figure, the tip of the acrylic transparent resin optical waveguide material 480 is kept straight and the surface mirror coat 478 is applied only to the outer portion. The shape may be easy to manufacture. In this case, it is more preferable to make the tip round to make it slippery.

Next, an example of a pen-type information reproducing apparatus using a light source integrated image sensor will be described (see FIG. 47).

That is, in this embodiment, the light source integrated image sensor 492 as described above with reference to FIG. 39 is used, and the rod lens (eg, SELFOC lens) as an image forming system is formed on the exposure surface thereof. 494 and a glass thin plate 496 are arranged and formed. Here, the thin glass plate 496 has a role of a protective glass with respect to an actual contact surface, and also has a role of providing a certain distance to make the illumination as flat as possible.

As described above, the light source integrated image sensor 4
By using 92, it is possible to reduce the size of the pen-type information reproducing device, and also to shorten it in the length direction.

Next, a pen type information reproducing apparatus for dealing with color multiplexed dot codes will be described.

FIG. 48A is a diagram showing the structure thereof. The touch sensor 466 as shown in FIG. 41B and the first and second touch sensors as shown in FIG. The polarizing filters 470 and 472 of FIG. Furthermore, the pen-type information reproducing apparatus of the present embodiment reads the color-multiplexed color code by synthesizing a plurality of dot codes of different colors as shown in FIG. 48B. Further, the color liquid crystal 498 controlled by the control unit 212 is arranged on the pupil plane of the lens 200.

Here, in order to describe the control method of the color liquid crystal 498 in the control section 212, first, an example of using a color multiplex dot code will be described.

For example, as shown in FIG.
The color multiple dot code 502 is arranged on the sheet 500, and corresponding to it, "Good Morning"
It is assumed that the characters are written and the index 504 and the index code 506 are arranged at a predetermined position, for example, the lower right. Then, when the color multi-dot code 502 is played back by this pen type information playback device, the pronunciation "Good morning" is output in Japanese, "Good morning" is pronounced in English, or "Good morning" is spoken in German. In order to select whether to pronounce “Guten Morgen”, as shown in (D) of the figure, the index code 506 arranged corresponding to the index 504 indicating the option is scanned and recognized, and, for example, Japanese is selected. After doing the color multiple dot code 502
The following explanation will be given for the purpose of making it possible to make a utterance such as "Good morning" by scanning.

First, as shown in (B) of the figure, a dot code for pronouncing in Japanese is generated and assigned as code 1 to red (R). Similarly, create a dot code to be pronounced in English as code 2, and then green (G)
, A dot code pronounced in German is created as code 3, and is assigned to blue (B). The color of the overlapping portion of each information is set as the color formed by the additive color of each color, and the color multiplex dot code 502 is applied to the sheet 500.
Record above. In this case, the portion where the colors do not overlap is recorded as a black dot. That is, as mentioned above, the dot code consists of a marker and a data dot, but the marker is black,
This means that the data dots are recorded in different colors by the color addition method. In this way, color multiple dot code 50
Recording at 2 means increasing the recording density.

Not limited to the three types of RGB colors, a plurality of different pieces of information may be assigned to colors of different narrow band wavelengths. More information such as types and 5 types can be multiplexed. As the color ink in that case, in addition to the conventional inks such as cyan, yellow, and magenta, it is possible to mix a dye (ink that reflects light of only a narrow band wavelength).

The index code 506 is arranged in the underlined portion of the index 504 indicated by characters or pictures so that the user can recognize and select it. Which color is selected for the printing. It is printed in black so that it can be read.

The color liquid crystal 498 is formed by adhering RGB light transmission mosaic filters to the pixels of the liquid crystal, and is for separating the information of each color of the color multiplex dot code 502. That is, the control unit 212 sets only the pixels corresponding to the color of the information selected by the scanning of the index code 506 into the transparent state.
Controlled by. Further, the liquid crystal may not have a mosaic shape, but the optical path may be divided into planes. that time,
It is preferable to make the divided area ratio of each color inversely proportional to the sensitivity of the pixel because the sensitivity for each color becomes uniform. That is, when the sensitivity of B is low, the area is made larger than that of other colors. The color liquid crystal may be placed on the light source side.

Next, the operation for reading the index code 506 and selecting a color to generate it in a desired language will be described with reference to the flowchart in FIG.

First, if green is selected (step S202) and the touch sensor 466 is pressed (step S204), the control section 212 controls the liquid crystal transmission part of the color liquid crystal 498 in accordance with the color selection. (Step S206). For example, since green is initially selected, only dots with a green filter are made transparent. Next, the light source 198 is controlled by the control unit 212, and the dot code is read by the image processing unit 460 (step S208). Then, the data processing unit 462 decodes the code (step S2
10) Recognize whether all the codes have been finished, that is, whether all the codes have been read (step S212). When the codes have been read, a sound is issued to notify that (step S21).
4). Next, the control unit 212 determines whether the index code 506 or the sound information (color multiple dot code 502) is read according to the decoding result (step S216). , The color indicated by the index code 506 is selected (step S218), and the process returns to step S204. If it is sound information, the data output unit 464 causes the sound output device 268 to reproduce sound (step S220).

After the sound reproduction in step S220, it is further determined whether or not the sound is repeatedly generated a predetermined number of times (step S222), and the number of times is preset by the repeat switch 467. If
The predetermined number of times is repeated and reproduced.

Of course, the number of repetitions may be one,
The number of times can be set by various switches and the like, and in addition to this, the number of times can be recorded in advance in the index code 506 or the dot code 502.

FIG. 15 shows the repeat reproduction here.
Alternatively, it is possible by repeatedly reading from the data memory unit 234 in FIG.

The image pickup section 204 includes a black and white one,
Generally, there is a color image pickup device in which a color mosaic filter is attached to the image pickup device section. In the above example, the monochrome image pickup unit is used. However, by using a color image pickup device and separating the colors in the image processing unit 460, it is possible to reproduce each color separately. In that case, the color liquid crystal 498 can be omitted.

FIG. 49B is a diagram showing the configuration of the image memory section of the image processing section 460 when a color image pickup device is used. That is, the signals coming from the color image pickup device are separated into respective colors by the color separation circuit 508 and stored in the memories 510A, 510B, 510C, which are selected by the multiplexer (MPX) 512 and the subsequent processing is performed. To do so.

Of the first and second polarization filters 470 and 472 for the purpose of preventing regular reflection, the second polarization filter 472 uses the same polarization filter in the polarizer portion of the color liquid crystal 498. Since it is known, it is possible to use it as well. Therefore, the color liquid crystal 498
The second polarization filter 472 can be omitted by combining with the other polarization filter. However, at that time, the angle in the horizontal plane of this color liquid crystal is
Must be rotated so as to cut the components in the same direction as the direction corresponding to the polarization filter 472, that is, in the same direction.

Further, as shown in FIG. 50A, the color liquid crystal 498 is removed, and the light source 198 is not a white light source but an RGB light source such as an LED as shown in FIG. 50B. Even, color multiple dot code 5
02 can be read. That is, in the case of dividing into RGB and the above three colors, among the RGB light sources 198, only the LED corresponding to red is turned on when reading the code 1 corresponding to red, and only the green LED corresponding to code 2 is read. For code 3, only the blue LED may be turned on to reproduce.

Also, instead of using RGB LEDs,
It is also possible to add a color filter to each part as a white light source to form a light source of each color.

As described above, by using the light sources of RGB different colors as the light source 198 and controlling the lighting of the light source of the color selected by the index code 506, the same effect as the configuration of FIG. 48A can be obtained. Obtainable. Furthermore, by having each of the light sources that emit light of a plurality of narrow band wavelengths, it is not necessary to have a color liquid crystal and its control circuit, and the size can be reduced at low cost. In particular, LEDs are narrow band,
For example, there is a light having a wavelength of about ± 27 nm, and if such a light is used, narrower band reproduction can be performed.

Next, a pen type information reproducing apparatus for stealth type dot codes will be described.

FIG. 51 (A) shows a dot data sticker 516 with a title on which an infrared light emitting paint dot code 514 as a stealth type dot code is printed.
This dot data sticker 516 is, for example, in a printing machine or a printer, prints a title, for example, in a normal color or black and white print, and prints a dot code below it by using an invisible paint. It was done. Of course, in this dot data sticker 516, since the dot code 514 is invisible, that is, transparent printing,
As shown in (B) of the figure, the dot code 514 may be printed over the title of visible information by using transparent ink. This printing is realized by, for example, in the case of an inkjet printer or the like, by adding an infrared light-emitting paint ink as a fifth ink to four inks of cyan, magenta, yellow, and black, and printing them in an overlapping manner. it can.

Note that FIG. 51A shows an example in which a title is printed in the margin of a stealth type dot code. Of course, the dot data sticker with a title is printed with a visible light dot code, A title may be printed in the margin.

As a pen type information reproducing apparatus for reproducing the infrared light emitting paint dot code 514 as such a stealth type dot code, for example, as shown in (C) of FIG. Since it is printed with luminescent paint, the infrared light emitting element 518 is used as the light source 198, and the infrared band bandpass optical filter 520 is arranged in front of the imaging unit 204.

That is, when the infrared light emitting element 518 irradiates the infrared light emitting paint dot code 514 with light in the infrared region, the light is reflected in the infrared region, that is, in a wavelength of a certain narrow wave band. In order to detect the intensity of the reflection by the imaging unit 204, the reflected light is guided by being separated from the visible light information through the infrared band bandpass optical filter 520.

Since it is possible to prepare several kinds of light emission bands of the paint used for printing the infrared light emitting paint dot code 514, for example, by changing the characteristics of the bandpass optical filter 520 little by little, the transparency can be obtained. Printing can also be multiplexed.

[0338]

As described in detail above, according to the present invention,
It is possible to provide a dot code capable of inexpensively recording a large amount of multimedia information including audio information, video information, digital code data, and the like, and capable of being repeatedly reproduced.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of a dot-coded audio information recording apparatus according to a first embodiment of the present invention.

FIG. 2A is a diagram showing a dot code recording format, and FIG. 2B is a diagram showing a usage state of the reproducing apparatus in the first embodiment.

FIG. 3 is a block configuration diagram of a playback device in the first embodiment.

4A and 4B are explanatory diagrams of manual scanning, and FIGS. 4C and 4D are explanatory diagrams of scan conversion.

FIG. 5A is a diagram for explaining data interpolation associated with scan conversion, and FIGS. 5B and 5C are diagrams showing examples of a recording medium.

FIG. 6 is an explanatory diagram of data string adjustment.

FIG. 7 is a diagram showing a configuration of a playback device according to a second embodiment.

FIG. 8 is a diagram showing a configuration of a playback device according to a third embodiment.

9A and 9B are external perspective views of a portable voice recorder.

FIG. 10 is a circuit configuration diagram of a portable voice recorder.

11A and 11B are diagrams showing an example of printing on a recording medium, and FIGS. 11C and 11D are perspective views showing the appearance of another example of a portable voice recorder.

12 is a flowchart of dot code printing processing in the voice recorder of FIG.

FIG. 13 is a block diagram of a multimedia information recording device.

FIG. 14 is a conceptual diagram of a dot code.

FIG. 15 is a block diagram of a multimedia information reproducing apparatus.

16 is a light source emission timing chart in the multimedia information reproducing apparatus of FIG.

FIG. 17 is a diagram showing another configuration example of the multimedia information reproducing apparatus.

18A is a diagram showing a dot code which is also applied to the reproducing apparatus of FIG. 3 for explaining the data string adjusting unit in the multimedia information reproducing apparatus of FIG. 15, and FIG. The figure showing the line-shaped marker of the dot code of
(C) is a diagram for explaining the scanning method, and (D) is a diagram for explaining the scan pitch of the image sensor.

FIG. 19 is a diagram showing an actual configuration of a data string adjustment unit.

20A to 20C are diagrams showing markers having dots for array direction detection, and FIG. 20D is a diagram showing still another configuration example of the multimedia information reproducing apparatus.

21A is an explanatory diagram of a block address, FIG.
Is a diagram showing a block configuration, (C) is a diagram showing an example of a marker pattern, and (D) is a diagram for explaining the magnification of the imaging system.

FIG. 22 is a block configuration diagram of a marker detection unit in the multimedia information reproducing apparatus.

23 is a processing flowchart of the marker determination unit in FIG. 22.

FIG. 24 is a processing flowchart of a marker area detection unit in FIG. 22.

25A is a diagram showing a marker area, FIG. 25B is a diagram showing a storage format of a table that stores the detected marker area, and FIGS. 25C and 25D are diagrams of FIG. FIG. 4 is a diagram showing a value obtained by accumulating each pixel in FIG.

FIG. 26 (A) is a process flowchart of the approximate center detection unit in FIG. 22, and FIG. 26 (B) is a flowchart of the center of gravity calculation subroutine in (A).

FIG. 27 is a block configuration diagram of an approximate center detection unit.

28A is a diagram showing an actual configuration of a dot code data block, FIG. 28B is a diagram showing another configuration, FIG.
(C) is a diagram for explaining another arrangement of the data inversion dots.

FIG. 29A is a diagram showing another example of the actual configuration of a dot code data block, and FIG. 29B is a diagram for explaining selection of adjacent markers.

[Fig. 30] Fig. 30 is a block configuration diagram of a data array direction detection unit in the multimedia information reproducing apparatus.

FIG. 31 is an operation flowchart of a data array direction detection unit.

32A is a flowchart of a neighboring marker selection subroutine in FIG. 31, and FIGS. 32B and 32C are diagrams for explaining neighboring marker selection.

33A is an explanatory diagram of direction detection, and FIG. 33B is a diagram for explaining a relationship between m and n in FIG.

FIG. 34 is an explanatory diagram of another method of direction detection.

35 (A) and (B) are respectively a block configuration diagram and an explanatory diagram of a block address detection, error determination, and accurate center detection unit in the multimedia information reproducing apparatus.

FIG. 36 is an operation flowchart of block address detection, error determination, and accurate center detection unit.

FIG. 37 (A) is a diagram for explaining the operation of the marker / address interpolator in the multimedia information reproducing apparatus, and FIG. 37 (B) is a block of the address controller in the multimedia information reproducing apparatus. It is a block diagram.

38A is a diagram for explaining another processing method of the marker determination unit, FIG. 38B is a diagram for explaining a marker determination formula, and FIG. 38C is a diagram for explaining marker alignment detection. Is.

FIG. 39 is a diagram showing a configuration of a light source integrated image sensor.

FIG. 40 is a one-chip IC using an XY address type image pickup unit.
It is a block configuration diagram of.

41A is a circuit configuration diagram of a pixel of an XY address type image pickup unit, and FIG. 41B is a diagram showing a configuration of a pen-type information reproducing device having a switch for controlling dot code fetching.

FIG. 42 is a three-dimensional IC using an XY address type image pickup unit.
It is a block configuration diagram of.

FIG. 43 is a diagram showing another configuration of a pen-type information reproducing device having a switch for controlling dot code import.

FIG. 44 (A) is a diagram showing a configuration of a pen-type information reproducing device capable of removing specular reflection, (B) is a diagram for explaining configurations of first and second polarization filters, and (C) is a second diagram. It is a figure which shows another structural example of the polarization filter of 2, and (D) is a figure which shows a structure of an electro-optic element shutter.

FIG. 45 is a diagram showing another configuration of a pen-type information reproducing device compatible with regular reflection removal.

46A is a diagram showing a configuration of a pen-type information reproducing device using a transparent resin optical waveguide material as a light source, and FIG. 46B is an enlarged view of a connecting portion between the optical waveguide material and the reproducing device housing. , (C) and (D) are diagrams showing the configuration of the tip portion of the optical waveguide material.

FIG. 47 is a diagram showing a configuration of a pen-type information reproducing device using an image sensor integrated with a light source.

48A is a diagram showing a configuration of a pen-type information reproducing apparatus compatible with color multiplexing, FIG. 48B is a diagram for explaining a color multiplex code, and FIG. FIG. 4D is a diagram for illustrating an index code.

FIG. 49A is an operation flowchart of the color-multiplexing-compatible pen-type information reproducing device, and FIG. 49B is a diagram showing a configuration of an image memory unit when a color imaging element is used.

FIG. 50A is a diagram showing another configuration of a color-multiplexing pen-type information reproducing device, and FIG. 50B is a diagram showing a configuration of a light source.

51A is a diagram showing a dot data sticker with a stealth-type dot code, FIG. 51B is a diagram showing a configuration of a pen-type information reproducing device compatible with the stealth-type dot code, and FIG. FIG. 6 is a diagram showing a dot data sticker in which a stealth type dot code is written in another mode.

[Explanation of symbols]

36, 170 dot code 36A, 36B manual scanning mark 38, 172, 304 block 38A, 174, 274, 310 marker 38B error correction code 38C audio data 38D x address data 38E y address data 38F error determination code 176, 272A, 306 block address 178 address error detection / correction data 180,314 data area 272B block address error correction data 276A line address 276B error detection data 278,316 dots 294A, 296A, 298A array direction detection dot 306A upper address code 306B Lower address code 308 Dummy marker 310A Circular black marker 310B White part of marker 312 Error detection code 312A Upper address CRC Over-de-312B lower address CRC code 364 data margin

─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Ken Mori 2-43-2, Hatagaya, Shibuya-ku, Tokyo Olympus Optical Co., Ltd. (56) Reference JP-A-3-127341 (JP, A) JP Flat 3-149683 (JP, A) Special Table Flat 6-502036 (JP, A) (58) Fields surveyed (Int.Cl. ⁷ , DB name) G06K 19/06

Claims (2)

(57) [Claims] 1. An optically readable dot code, wherein a plurality of blocks are arranged so as to be adjacent to each other, and each of the blocks includes a non-modulation region including a marker for recognizing the block, Each of the bit values of the data subjected to the modulation processing so as to be structurally distinguishable from the above marker is composed of a modulation area including a data dot pattern composed of a plurality of dots arranged in dots corresponding to the value. The non-modulated area further includes a block address pattern for indicating an address of the block. 2. An optically readable dot code, wherein a plurality of blocks are arranged so as to be adjacent to each other, and each of the blocks includes a non-modulation region including a marker for recognizing the block, Each of the bit values of the data subjected to the modulation processing so as to be structurally distinguishable from the above marker is composed of a modulation area including a data dot pattern composed of a plurality of dots arranged in dots corresponding to the value. The non-modulation area further includes a block address pattern for indicating an address of the block, and a dot pattern for pattern matching for determining a read position of each dot in the data dot pattern. And the dot code.